Download document 8137627

Journal of Heredity 2014:105(Special Issue):771–781
doi:10.1093/jhered/esu043
© The American Genetic Association. 2014. All rights reserved.
For permissions, please e-mail: [email protected]
Comparative Transcriptomics of
Maturity-Associated Color Change in
Hawaiian Spiders
KRISTINA M. YIM, MICHAEL S. BREWER, CRAIG T. MILLER, AND ROSEMARY G. GILLESPIE
From the Department of Molecular and Cell Biology, University of California, 142 Life Sciences Addition #3200, Berkeley,
CA 94720-3200 (Yim and Miller); and the Department of Environmental Science, Policy, and Management, University of
California, Berkeley, CA 94720-3114 (Brewer and Gillespie).
Abstract
ciation. This study is part of an effort to investigate the molecular genetic underpinnings of adaptive radiation in Hawaiian
spiders (genus Tetragnatha). This radiation is found throughout the Hawaiian Islands, showing a common pattern of evolutionary progression from older to younger islands. Moreover, the species are characterized by repeated evolution of similar
ecomorphs that can be recognized on the basis of color—Green, Maroon, Large Brown, and Small Brown. However, 2 species (including T. kauaiensis
developmental period. The current study focuses on the age-associated color change in the early stages of the radiation to
T. kauaiensis and T. perreirai
polyphenism and associated switch to separate monophenic ecomorphs. These results provide critical groundwork that will
allow us to advance our understanding of the genomic elements associated with adaptive radiations.
Subject areas: Molecular adaptation and selection
Key words: adaptive radiation, color evolution, dN/dS, polyphenism, spider, transcriptomics
that result in clades with broad phenotypic diversity but
without comparable levels of genetic divergence (Givnish
). The ecological disparity indicates that
small changes of large effect are likely of particular importance in adaptive radiations (
). This has led to
the suggestion that suites of characters involved in adaptive
radiation may be correlated (
). Indeed,
parallel evolution of similar ecomorphological attributes is
a common feature in adaptive radiation and has been well
)
), Anolis lizards of
the Caribbean (
), sticklebacks of postglacial lakes
(
), Mandarina snails of the Bonin Islands (Chiba
), and Hawaiian spiders (
), among others
(
). In each of these radiations, the number of
form—or “ecomorph”—generally represented by multiple
species, each occurring in a different geographic area.
vergent evolution involves deployment of similar genetic
pathways (
;
), the link with development has been demonstrated only recently. In particular,
polyphenism is a special case of phenotypic plasticity in which
the same genotype can lead to 2 or more phenotypes as a
the social or seasonal environment (
); the
phenomenon is known to be associated with adaptive radiation (
). In particular, developmental
polyphenism—an organism’s ability to change its phenotype
(or morph) during its lifetime—is being increasingly recognized as major factor in ecological speciation (Fitzpatrick
). Such variability within an individual may provide a
mechanism for evolutionary change by making it easier for
771
Downloaded from http://jhered.oxfordjournals.org/ at East Carolina University on August 27, 2014
Address correspondence to Kristina M. Yim at the address above, or e-mail: [email protected].
Journal of Heredity
; Moczek
). The most detailed studies on the role of developmental polyphenism in facilitating adaptive differentiation
have been performed on fish—in particular, sticklebacks and
cichlids (
;
; McGuigan
;
lebacks, ontogenetic studies suggest that individuals of anadromous species are limnetic when young, becoming more
(
). Most species of Tetragnatha are uniform
in appearance—dull brown or olive in color, long jaws
in adulthood, long first and second legs, and an elongate
opisthosoma (
)—and overall are characterized by
fairly homogeneous behavior and ecological affinity as they
generally construct a flimsy orb web with an open center
and build the web over or near water (
). In
the Hawaiian Islands, the remarkable adaptive radiation of
Tetragnatha
ecological affinities that are not seen in the genus elsewhere
in its range (
). In particular, the “spiny-leg”
Hawaiian Tetragnatha
webs, adopting a wandering lifestyle instead. These species
appearance, particularly color (Green, Maroon, Large Brown,
or Small Brown), and the substrate upon which they find refuge during the day: Green (refuge on green leaves), Maroon
(on mosses), Small Brown (on twigs), or Large Brown (on
branches;
;
nocturnal behavior of the spiders and their very limited
visual capacity, diurnal predation is the most likely selective
pressure responsible for the close color matching (
); the most likely predators are honeycreepers, for which spiders can form an important component of the diet (
species appear to have arisen through a combination of
772
case of T. perreirai). These species, T. perreirai and T. kauaiensis
(Figures 1 and 2), form the focus of the current study.
Given that both specimens used belonged to the Maroon
ecomorph, the differences between them are unlikely to be due
to differences in pigmentation. The islands where these 2 species
occur are adjacent, and the habitats are very similar (Metrosiderosdominated wet forest, with the spiders occurring at similar elevations), so environmental differences are unlikely to lead to major
loci should be due either to: 1) Differences in the genetic environment. Colonization of a new island will inevitably be associated with a population bottleneck and associated change in gene
speciation and diversification (
).
2) Selection acting on the loci involved in changing or not changing color, the polyphenic T. kauaiensis (on the older island) showin the more derived T. perreirai (on a younger island; Brewer et al.
forthcoming;
;
).
In reality, each of these factors is likely to lead to differences in selective pressures. By identifying the genes that are
that arise in different environments may depend on the initial
appearance of these phenotypes in a phenotypically plastic,
esis,
). The current system provides the
radiation because the Hawaiian islands are arranged chronologically, the spiny-leg clade of Tetragnatha spiders have colonized and radiated within each current high islands, and the
tendency to change from 1 ecomorph at an early adult stage
to another in a later adult stage, is largely confined to the
use this natural replication coupled with a gradation in polyphenism, to understand the dynamics of adaptive radiations
developmental polyphenism in a radiation of Hawaiian spiders and determining its role in adaptive radiation, the current
Downloaded from http://jhered.oxfordjournals.org/ at East Carolina University on August 27, 2014
limnetic and benthic forms) may then arise by alteration
tive traits (
). Similar effects have been
found in other radiations, such as Geospizine finches of the
Galapagos in which selection on developmental plasticity in
beak shape may have played a role in species diversification
(
).
trolled color polymorphism (
).
However, documentation of polyphenic variability is limited
ing to environmental conditions (
;
) or diet (
recent work has shown that certain spiders in the Hawaiian
Islands, in particular the more basal representative within an
adaptive radiation of long-jawed spiders (genus Tetragnatha),
Brewer et al. forthcoming;
).
The long-jawed orb-weaving spider genus Tetragnatha
unidirectional interisland dispersal and intraisland or intravolcano speciation (
,
).
In addition to showing repeated evolution of ecomorphs,
Hawaiian spiny-leg Tetragnatha display developmental polyphenism, which is most pronounced on the older islands
(Kauai and Oahu), such that 2 species from the clade on the
oldest islands change from 1 color-associated ecomorph to
another over the course of development (
;
). On the younger islands, the polyphenism is
out their lifecycle (
). Tetragnatha kauaiensis from
Kauai, the oldest of the currently high islands, undergoes the
most pronounced developmental color change from Green
to Maroon in early to late maturity, as does T. polychromata
from Oahu (
). In contrast, T. perreirai from Oahu,
:JNFUøBMr5SBOTDSJQUPNJDTPG$PMPS$IBOHFJO)BXBJJBO4QJEFST
Downloaded from http://jhered.oxfordjournals.org/ at East Carolina University on August 27, 2014
Figure 1. Phylogeny of spiny leg Hawaiian Tetragnatha, showing T. perreirai and T. kauaiensis
allozymes (
). Volcano age of the collection sites on Kauai and Oahu are indicated in millions of years (myrs).
Figure 2. Tetragnatha kauaiensis, same individual (mature) at (A) 3-days postmaturity (Green, color unchanged since hatching);
(B
C) T. perreirai (Maroon).
study is a preliminary effort to generate genomic tools for this
nonmodel system. Thus, the objectives of the current work
Materials and Methods
Maroon T. perreirai and the Maroon morph of the polyphenic
T. kauaiensis, 2) annotate relevant loci, 3) detect signatures of
Live specimens of the monophenic Maroon T. perreirai
and the Maroon ecomorph of the polyphenic T. kauaiensis were collected from the Hawaiian Islands, snap-frozen,
the genetic basis of developmental polyphenism and understanding its role in adaptive radiations and rapid speciation.
The following specimens were used: 1 T. perreirai collected
from Oahu (Mount Ka’ala) and 1 Maroon ecomorph of
Total RNA Extraction from Whole Bodies
773
Journal of Heredity
T. kauaiensis
each gene in the transcriptomes using the TransDecoder utility (http://transdecoder.sourceforge.net/
-
that has been shown to perform well in small arthropods
). The entire
Data Set Construction and Annotations
were then sorted into respective files for each pairwise com-
-
protocols.
program (http://www.bioinformatics.babraham.ac.uk/proof the resulting paired-end reads. The raw reads generated by
(
): nucleotides with
-
trimmed to remove primer artifacts. Individual paired-end
files were then resynchronized, removing any paired-end
Transcriptome Assembly, Contaminant Screening, and
Open Reading Frame Prediction
Transcriptomes of each specimen were assembled from
the preprocessed reads using the Trinity pipeline (http://
;
), using a
ratio tests and d /dS ratios were used to identify the best
model for each gene: a null hypothesis of neutral evolution (P
) or natural selection (positive selection
[d /dS > 1] or stabilizing selection [d /dS
). P-values
from the likelihood ratio tests were corrected for multiple
comparisons using the false discovery rate (Benjamini and
http://
;
).
Contigs were annotated using the programs Blast2GO
(http://www.blast2go.com/;
), which
retrieves gene ontology (GO) terms and selects functional
http://ab.inf.uni-tuebingen.
de/software/megan/;
), which utilizes
contigs to broad functional categories and looking at the
distribution of gene functions at the transcriptome level,
annotations provide a biological foundation for globally characterizing the transcriptome. In fulfillment of data archiving
guidelines (
), we have deposited the primary data
underlying these analyses with Dryad.
k-mer coverage of 2. Trinity was designed specifically for
Results
refine de Bruijn graph contigs. This iterative refinement has
Quality Assessment and Preprocessing of Illumina
RNA-Seq Reads
tive splicing (
). The assembled contigs
). High-confidence open reading frames
774
T. perreirai
for the Maroon morph of T. kauaiensis
-
retained for T. perreirai
retained for T. kauaiensis. The FastQC program was used to
assess GC content (Figure 3
across all bases before and after preprocessing (
).
Downloaded from http://jhered.oxfordjournals.org/ at East Carolina University on August 27, 2014
Illumina RNA Sequencing and Preprocessing
http://mafft.cbrc.jp/alignment/software/;
then transposed onto this alignment using pal2nal.pl (http://
www.bork.embl.de/pal2nal/;
), and gaps
http://www.
tcoffee.org/;
).
http://abacus.gene.ucl.ac.uk/software/paml.html;
) was used to analyze each nucleotide alignment for signatures of selection by estimating
the ratio of synonymous to nonsynonymous substitutions
(d /dS) and calculating likelihood values for different selec-
:JNFUøBMr5SBOTDSJQUPNJDTPG$PMPS$IBOHFJO)BXBJJBO4QJEFST
Transcriptome Assembly, Contaminant Screening, and
Predicted ORFs
rier activity,” and “rhythmic process.”
T. per-
Putative Orthologs and Analyses
T. perreirai and T. kauaiensis (Maroon)
reirai
T. kauaiensis from
Table 1).
for mitofor other contaminants). Seventy-three homologous
contigs were removed for T. perreirai
T. kauaiensis
Table 1).
) and likelihood values for difd /dS ratios (shown in
ferent selection models for each orthologous gene. These data
were used to choose the best selection model for each gene:
32 genes showed signatures of positive selection (d /dS > 1),
d /dS
P
signatures of positive selection are described in Table 2.
Discussion
T. perreirai
T. kauaiensis.
Functional and Taxonomic Annotations
T. perreirai
T. kauaiensis. The mean size of assem-
T. perreirai. For T. kauaiensis
nomic annotations. Functional annotations were categorized
gies: cellular component, molecular function, and biological
process (
and “metabolic process.” In comparison, a very small percent
for T. perreirai
for T. kauaiensis.
assembly is dependent on k-mer length, the user-defined
k-mer length may
increase the total number of contigs assembled, but these
775
Downloaded from http://jhered.oxfordjournals.org/ at East Carolina University on August 27, 2014
Figure 3.
Tetragnatha perreirai and the Maroon
morph of T. kauaiensis
corresponds to the overall GC content of the underlying transcriptome. The 2 individuals were found to have similar overall GC
T. perreirai
T. kauaiensis (Maroon).
Journal of Heredity
Downloaded from http://jhered.oxfordjournals.org/ at East Carolina University on August 27, 2014
Figure 4.
Tetragnatha perreirai and T. kauaiensis
y
Table 1
Transcriptome assembly and contaminant screening summary
Tetragnatha perreirai
)
screening (percent of total contigs)
776
)
)
1
Tetragnatha
kauaiensis (Maroon)
1
:JNFUøBMr5SBOTDSJQUPNJDTPG$PMPS$IBOHFJO)BXBJJBO4QJEFST
Downloaded from http://jhered.oxfordjournals.org/ at East Carolina University on August 27, 2014
Figure 5. Contig length distribution of Trinity assembly for Tetragnatha kauaiensis (Maroon) and T. perreirai. Transcriptomes for
T. perreirai and T. kauaiensis
Figure 6.
Tetragnatha perreirai and T. kauaiensis (Maroon).
component, molecular function, and biological process.
contigs may be highly fragmented and increase the possibility
of mis-assembly. Because reads are decomposed into k-mers,
loss of information is possible. Paralogs with high levels of
cannot be distinguished from 1 another due to short read
length and lack of a reference genome.
T. perreirai
T. kauaiensis
functionally annotated, respectively, using Blast2GO. GO
annotations comprehensively describe properties of specific genes and their products, and can be used to predict
777
Journal of Heredity
Table 2
Genes showing signatures of positive selection
Lengtha
Test #
321
dN/dS
LRb
P value
Functionc
Taxond
Binding
Mitochondrial inner membrane
Intracellular membrane-bounded organelle
Intracellular
Metazoa
Panarthropoda
331
Panarthropoda
111
Panarthropoda
Panarthropoda
Binding
333
aLength
Bilateria
Polytene chromosome puff
of alignment.
ratio from the likelihood ratio test of neutral selection versus positive or stabilizing selection.
bLikelihood
c
d
the physiological role of each gene. Contigs were assigned to
a broad range of GO categories, indicating that data generand functions. Many genes were assigned to the GO categories “cell,” “binding,” “cellular process,” and “metabolic
tated for T. perreirai and T. kauaiensis, respectively. For both
annotation fell under either “Panarthropoda” or “Bilateria.”
There was also a small percent of genes in both transcripidentify and remove contaminants that were missed by the
initial contaminant screening and removal.
and their functions remain unknown or not well characterized.
These results are also indicative of the limitations of inferring relevant functions of genes from de novo transcriptome
assemblies for species with very limited genomic information.
ferent transcriptomes are assumed to be orthologous if they
find each other as the best hit in the other transcriptome.
The majority of orthologs showed signatures of stabilizing selection. Thirty-two orthologs were identified as being
under positive selection and may be involved in the loss of
an ancestral developmental polyphenism and the associated
switch to separate monophenic ecomorphs. However, many
of these genes did not receive a definitive annotation, making
it difficult to perform functional prediction and classification
for putative orthologs. Beyond being associated with color
changing, these loci may contribute to a number of other
traits that have diverged as a result of life-history differences,
new environments (the species inhabit different, though
adjacent, islands of the Hawaiian archipelago).
778
Downloaded from http://jhered.oxfordjournals.org/ at East Carolina University on August 27, 2014
Panarthropoda
212
:JNFUøBMr5SBOTDSJQUPNJDTPG$PMPS$IBOHFJO)BXBJJBO4QJEFST
Coding Loci Evolution and Signatures of Selection
The current study showed evidence of selection on some
transcriptomes. Given that both specimens belonged to the
Maroon ecomorph, the differences between them are unlikely
to be due to differences in pigmentation. T. kauaiensis occurs
on Kauai, and T. perreirai on Oahu. These islands are adjacent, and the habitats are very similar (Metrosideros-dominated
wet forest, with the spiders occurring at similar elevations),
so environmental differences are unlikely to lead to major
d /dS > 1).
Annotation and Limitations of the Current Study
selection in ecological divergence and speciation within a radiation of Hawaiian spiders, we cannot yet provide definitive
resources for spiders (Brewer et al. forthcoming): There is no
reference spider genome, though there are a number of efforts
that promise these resources shortly. In particular, recent efforts
http://www.arthropodgenomes.org/
1) Differences in the genetic environment. Colonization of
a new island will inevitably be associated with a population
kind of effect has been argued as initiating speciation and
diversification (
). 2) Selection
acting on the polyphenism (of T. kauaiensis) that has lead to
monophenism in T. perreirai
shift from polyphenism in the more “ancestral” (on the oldest
island) T. kauaiensis
in the more derived T. perreirai (on a younger island; Gillespie
;
) is as follows: Given the presumed role of
color in crypticity, it is critical that the spiders select approthat the spiders do not select substrates based on color per
and lack visual acuity (
); presumably they are
of substrates, but may be less precise in their selection of
appropriate substrate than those that maintain a single color.
In reality, each of these factors is likely to lead to differences in selective pressures. By identifying the genes that are
responsible for that selection.
tissues transcriptomes. Moreover, there are a number of efforts
that are focusing in particular on uncovering the genetics and
molecular biology of silk production (e.g., Garb and Hayashi
;
;
; Garb et al.
;
) and the sophisticated combinatorial
evolution of spider venoms (
; Zobel-Thropp
;
). Moreover, in Hawaiian
spiders, there has been recent focus on the genomics of pigand assembled the transcriptomes of 2 spiders in the family
Theridiidae that have a very comprehensive inventory of the
putative gene content (
) and have a large
amount of genomic data for the Hawaiian Happy Face spider
(Theridion grallator)—including Illumina and PacBio data, which
are currently being assembled.
Future Directions
The combination of transcriptome assembly and characteriframework for genomic and transcriptomic studies to identify the molecular genetic underpinnings of color polymorphism, maturity-associated color change, and the divergence
779
Downloaded from http://jhered.oxfordjournals.org/ at East Carolina University on August 27, 2014
Figure 7. Pairwise d /dS values for putative orthologs. d /dS
shown, R2
Journal of Heredity
of species in the spiny leg clade of Hawaiian Tetragnatha.
identify the genes specifically involved in adaptive differentiation, genomic resources are accumulating. In particular, recent
work has shown the presence of all ommochrome pathway
genes in theridiid spiders (
), suggesting
that the color change in T. kauaiensis must be due to 1) syn-
state of ommochromes (
). The transcriptome data we provide represents some of the first for
-
snails of the genus Mandarina
Blast2GO: a universal tool for annotation, visualization and analysis in func-
novo characterization of the gene-rich transcriptomes of two color-polymorphic spiders, Theridion grallator and T. californicum
-
tifying the molecular basis for adaptive evolutionary change,
and potentially the triggers for adaptive radiation itself.
-
-
Acknowledgments
Speciation Continuum symposium and inviting this contribution to the jourorb weaving spider Tetragnatha elongata (Araneae, Tetragnathidae
spider Theridion grallator (Araneae, Theridiidae
References
Tetragnatha: I. Spiny leg
-
Blueprint for a high-performance biomaterial: full-length spider dragline silk
-
the primary constituent of dragline silk, in widow spiders (Latrodectus:
-
proposed nomenclature of the gene family that includes sphingomyelinase
Dolichognatha and
habitats, morphology and selective pressures: developmental polyphenism in
-
780
Tetragnatha
Downloaded from http://jhered.oxfordjournals.org/ at East Carolina University on August 27, 2014
Funding
:JNFUøBMr5SBOTDSJQUPNJDTPG$PMPS$IBOHFJO)BXBJJBO4QJEFST
-
-
Cryptic genetic variation and body size evolution in threespine stickleback.
tern to determining process and mechanism of evolution. Science.
-
-
http://
research.amnh.org/iz/spiders/catalog
Received April 21, 2014; First decision May 30, 2014;
Accepted June 10, 2014
Corresponding Editor: Sean Mullen
781
Downloaded from http://jhered.oxfordjournals.org/ at East Carolina University on August 27, 2014
stem” model of evolution: ancestral plasticity, genetic accommodation, and mor-