Article

Evidence for Faster X Chromosome Evolution in Spiders

Abstract and Figures

In species with chromosomal sex determination, X chromosomes are predicted to evolve faster than autosomes because of positive selection on recessive alleles or weak purifying selection. We investigated X chromosome evolution in Stegodyphus spiders that differ in mating system, sex-ratio, and population dynamics. We assigned scaffolds to X chromosomes and autosomes using a novel method based on flow cytometry of sperm cells and reduced representation sequencing. We estimated coding substitution patterns (dN/dS) in a subsocial outcrossing species (S. africanus) and its social inbreeding and female biased sister species (S. mimosarum), and found evidence for faster-X evolution in both species. X chromosome to autosome diversity (piX/piA) ratios were estimated in multiple populations. The average piX/piA estimates of S. africanus (0.57 (95% CI: 0.55-0.60)), was lower than the neutral expectation of 0.75, consistent with more hitchhiking events on X-linked loci and/or a lower X chromosome mutation rate, and we provide evidence in support of both. The social species S. mimosarum has a significantly higher piX/piA ratio (0.72 (95% CI: 0.65-0.79) in agreement with its female biased sex ratio. Stegodyphus mimosarum also have different piX/piA estimates among populations, which we interpret as evidence for recurrent founder events. Simulations show that recurrent founder events are expected to decrease the piX/piA estimates in S. mimosarum, thus underestimating the true effect of female biased sex ratios. Finally, we found lower synonymous divergence on X chromosomes in both species, and the male-to-female substitution ratio to be higher than 1, indicating a higher mutation rate in males.
Content may be subject to copyright.
Evidence for Faster X Chromosome Evolution in Spiders
Jesper Bechsgaard,
†,1
Mads Fristrup Schou,
†,1,2
Bram Vanthournout,
1,3
Frederik Hendrickx,
4,5
Bjarne Knudsen,
6
Virginia Settepani,
1
Mikkel Heide Schierup,
1,7
and Trine Bilde*
,1
1
Department of Bioscience, Aarhus University, Aarhus C, Denmark
2
Department of Biology, Lund University, SE-223 62 Lund, Sweden
3
Evolution and Optics of Nanostructure Group (EON), Biology Department, Ghent University, Ghent, Belgium
4
Royal Belgian Institute of Natural Sciences, Brussels, Belgium
5
Terrestrial Ecology Unit (TEREC), Biology Department, Ghent University, Ghent, Belgium
6
Qiagen Bioinformatics, Aarhus, Denmark
7
Bioinformatics Research Centre (BiRC), Aarhus University, Aarhus C, Denmark
These authors contributed equally to this work.
*Corresponding author: E-mail: trine.bilde@bios.au.dk.
Associate editor: Amanda Larracuente
Abstract
In species with chromosomal sex determination, X chromosomes are predicted to evolve faster than autosomes because
of positive selection on recessive alleles or weak purifying selection. We investigated X chromosome evolution in
Stegodyphus spiders that differ in mating system, sex ratio, and population dynamics. We assigned scaffolds to X
chromosomes and autosomes using a novel method based on flow cytometry of sperm cells and reduced representation
sequencing. We estimated coding substitution patterns (dN/dS) in a subsocial outcrossing species (S. africanus)andits
social inbreeding and female-biased sister species (S. mimosarum), and found evidence for faster-X evolution in both
species. X chromosome-to-autosome diversity (piX/piA) ratios were estimated in multiple populations. The average piX/
piA estimates of S. africanus (0.57 [95% CI: 0.55–0.60]) was lower than the neutral expectation of 0.75, consistent with
more hitchhiking events on X-linked loci and/or a lower X chromosome mutation rate, and we provide evidence in
support of both. The social species S. mimosarum has a significantly higher piX/piA ratio (0.72 [95% CI: 0.65–0.79]) in
agreement with its female-biased sex ratio. Stegodyphus mimosarum also have different piX/piA estimates among
populations, which we interpret as evidence for recurrent founder events. Simulations show that recurrent founder
events are expected to decrease the piX/piA estimates in S. mimosarum, thus underestimating thetrueeffectoffemale-
biased sex ratios. Finally, we found lower synonymous divergence on X chromosomes in both species, and the male-to-
female substitution ratio to be higher than 1, indicating a higher mutation rate in males.
Key words: sex chromosome, social spider, faster-X, female bias.
Introduction
In many species with chromosomal sex determination sys-
tems, males are hemizygous for the sex chromosomes and
loci harbored on sex chromosomes may therefore evolve
faster than similar loci on the autosomes, an effect termed
“faster-X.” Faster-X is caused by an elevated nonsynonymous
substitution rate on the X chromosome if 1) new advanta-
geous mutations are, on average, at least partially recessive,
because recessive alleles are exposed to selection in the hemi-
zygous state; and/or 2) if selectionislessefficientagainst
deleterious mutations on X chromosomes because of the
smaller effective population size of the X chromosomes com-
paredwiththeautosomes(Charlesworth et al. 1987;Vicoso
and Charlesworth 2006;Hedrick 2007;Ellegren 2009;Wright
et al. 2015). Empirical data provide conflicting conclusions on
the existence and generality of “faster-X” evolution of X chro-
mosomes (see supplementary table 1,Supplementary
Material online). For example, consistent evidence of faster-
X evolution comes from studies on mammals (Lu and Wu
2005;Torgerson and Singh 2006;Carneiro et al. 2012;Hvilsom
et al. 2012;Xu et al. 2012)andbirds(faster-Z)(Mank et al.
2007,2010;Wright et al. 2015). Conversely, in Drosophilids a
number of studies provide inconsistent evidence for faster-X
(Betancourt et al. 2002;Counterman et al. 2004;Thornton
et al. 2006;Hu et al. 2013), potentially due to low power
(Charlesworth et al. 2018), and the same is true for the few
other insect species studied (Jaquiery et al. 2012,2018;
Sackton et al. 2014;Rousselle et al. 2016). Different explana-
tions proposed for this inconsistency includes both selective
and ecological forces. A useful approach to study the forces
causing variation in the evolution of X chromosomes is the
study of closely related species that differ in traits predicted to
affect X chromosome to autosome (X/A) divergence. For ex-
ample, differences in life history traits and mating system,
such as age at sexual maturity and polyandry, are proposed
to underlie differences in X/A divergence of silent and coding
sites among four primate species (Xu et al. 2012), but the
number of such comparative studies are still very limited.
Article
ßThe Author(s) 2019. Published by Oxford University Press on behalf of the Society for Molecular Biology and Evolution.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/
licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is
properly cited. Open Access
Mol. Biol. Evol. 36(6):1281–1293 doi:10.1093/molbev/msz074 Advance Access publication March 26, 2019 1281
Downloaded from https://academic.oup.com/mbe/article-abstract/36/6/1281/5420164 by Ghent University user on 11 March 2020
Populations always carry fewer X chromosomes than auto-
somes, and under the neutral expectation this leads to rela-
tively fewer recombination events and higher rates of drift,
which in turn decreases nucleotide diversity on X-chromo-
somes (Ellegren 2009;Ellegren and Galtier 2016). Based solely
on the relative numbers of X chromosomes to autosomes in
species with equal sex ratio, the diversity of the X chromo-
some is predicted to be 0.75 of that of the autosomes
(Ellegren 2009). However, because the relative diversity of X
chromosomes to autosomes (piX/piA) is influenced by differ-
ent evolutionary forces including different mutation rates on
X chromosomes and autosomes, population size fluctuations,
breeding system, and recombination rate, piX/piA may devi-
ate from 0.75 (Ellegren 2009). Disentangling the relative influ-
ence of these forces on the diversity on X chromosomes
versus autosomes is important for our understanding of
how molecular evolution shapes genomes (Charlesworth
et al. 1987;Miyata et al. 1987;Ellegren 2007,2009;Pool and
Nielsen 2007,2008). Deviation from the null expectation of X
chromosome diversity of 0.75 of autosomal diversity is often
used to infer evolutionary history. For example, piX/piA esti-
mates of <0.75 in non-African populations of both humans
and Drosophila were interpreted to be caused by founder
events associated with “out of Africa” dispersal (Pool et al.
2012;Arbiza et al. 2014). In Drosophila, different sex ratios in
African (unbiased) and European (male biased) populations
were inferred by contrasting polymorphism data from X chro-
mosomes and autosomes (Hutter et al. 2007).
Comparisons of closely related species have proven useful
for elucidating genetic consequences of biological differences,
because of their recently shared evolutionary history (Cutter
et al. 2008;Guo et al. 2009;Settepani et al. 2017). Here, we
present a study of two sister species with contrasting mating
systems from the spider genus Stegodyphus: the subsocial
outcrossing species S. africanus and its social inbreeding sister
species S. mimosarum, with the subsocial outcrossing S. line-
atus as an outgroup (fig. 1)(Johannesen et al. 2007;Settepani
et al. 2016). The aim is to investigate how differences in biol-
ogy and mating system may influence the evolution of auto-
somes and sex chromosomes. Stegodyphus spiders have an X0
sex determining system, where females have two copies of
two X chromosomes (X
1
X
2
/X
1
X
2
) and males have one copy of
the two X chromosomes (X
1
X
2
/0) (Forman M, personal
communication). Differences in their degree of sociality and
mating system, and associated life histories and population
dynamics, are expected to influence substitution and diversity
patternsofXchromosomesandautosomesdifferently:The
subsocial outbreeding S. africanus has an equal primary sex
ratio (Vanthournout et al. 2018), and populations are
expected to be relatively stable in sizes and existence over
evolutionary time (Lubin and Bilde 2007;Settepani et al.
2017). In contrast, the social obligatory inbreeding S. mimo-
sarum shows a highly female-biased primary sex ratio (Lubin
and Bilde 2007), caused by male production of a higher pro-
portion of X
1
X
2
-containing sperm cells than sperm cells
without X chromosomes (Vanthournout et al. 2018).
Furthermore, empirical data suggest that population extinc-
tion rates in social Stegodyphus species such as S. mimosarum
are high (Crouch and Lubin 2001;Bilde et al. 2007), implying a
high rate of population colonization (Bilde et al. 2007), a
pattern supported by recent population genomic analyses
(Settepani et al. 2017). Differences in sex ratio influence the
relative effective population sizes of X chromosomes and
autosomes. If sex ratio is female biased, as in the social S.
mimosarum, the effective population size of X chromosomes
approaches that of autosomes, predicting similar evolution-
ary dynamics on X and A. This effect will be counter-acted if
the operational sex ratio is less female biased because of fe-
male reproductive skew and cooperative breeding (Lubin and
Bilde 2007;SalomonandLubin2007;Junghanns et al. 2017).
Population size fluctuation is also an important factor, as
population size reduction is predicted to more rapidly reduce
X chromosome diversity relative to autosome diversity, while
population growth is predicted to more rapidly elevate X
chromosome diversity relative to autosome diversity. This is
because population fluctuations influence effective popula-
tion sizes of X chromosomes and autosomes differently, for
example, Ne
X
will experience a relatively faster decline than
Ne
A
under a population reduction (bottleneck) (Pool and
Nielsen 2007).
We developed a new, cost-effective, and highly efficient
approach to sort scaffolds from the S. mimosarum genome
sequence (Sanggaard et al. 2014) into X chromosomes and
autosomes using a combination of flow cytometry and re-
duced representation (RAD) sequencing. Subsequently, we
applied transcriptome sequencing to generate estimates of
X chromosome and autosome substitution patterns (dN/dS),
and RAD sequencing to determine genetic diversity (pi) of X
chromosomes and autosomes. With this data, we assessed
theoretical predictions of how differences in sex ratio and
population size dynamics affect X chromosome relative to
AB
S. sarasino rum
S. pacificus
S. dufuori
S. bicolor
S. dumicola
S. tentoriicola
S. mimosarum
S. africanu s
S. tibialis
S. lineatus
FIG.1. Study system. (a) Phylogeny of the spider genus Stegodyphus
(modified after Settepani et al. 2016). Social species are underlined, and
the species included in this study are boxed in gray. (b) Geographic
location of sampled S. mimosarum populations (MAH, SAK, TANA,
WEE, PON) and of S. africanus populations (WRF, PON, KRU).
Bechsgaard et al. .doi:10.1093/molbev/msz074 MBE
1282
Downloaded from https://academic.oup.com/mbe/article-abstract/36/6/1281/5420164 by Ghent University user on 11 March 2020
autosome evolution in two closely related Stegodyphus spe-
cies (S. africanus and S. mimosarum)(fig. 1).
Results
Assigning Scaffolds to the X Chromosome
We were able to isolate the nuclei from sperm cells extracted
from an S. mimosarum male pedipalp, and separate the nuclei
with and without X chromosomes using flow cytometry.
Using RAD sequencing of the nuclei, we obtained more
than 1 million reads after quality filtering from each sample
that were subsequently mapped to the reference genome of
S. mimosarum.Figure 2 shows a density distribution of the
number of reads from the sample without X chromosomes
(“Sample 0”) divided by the total number of reads from both
samples (“Sample 0 þSample X
1
X
2
)mappedtoeachscaf-
fold (see Materials and Methods for details). The distribution
is bimodal (fig. 2 and supplementary fig. 1,Supplementary
Material online). The major peak close to P
0
¼0.5 shows that
most scaffolds have a similar coverage in “Sample 0” and
“Sample X
1
X
2
,” while the minor peak around P
0
¼0.119
constitutes scaffolds with much lower coverage in “Sample
0” compared with “Sample X
1
X
2
,” suggesting that they are
placed on the X chromosomes. We used a threshold of P
0
<0.238 to select scaffolds that we assign to the X
chromosomes and P
0
>0.3 as a threshold for a scaffold to
be considered autosomal, with both thresholds correspond-
ing to a FDR of 2.5% (supplementary fig. 1 and table 2,
Supplementary Material online). In this way, we obtained
450 X chromosome scaffolds for downstream analyses
(a list assigning scaffolds to X chromosome and autosome
scaffolds can be found in supplementary table 3,
Supplementary Material online).
Characterizing the X chromosome scaffolds using the
S. mimosarum reference genome, we found 2,132 X-
linked genes across 246.7 Mb or one gene per
115,725 bp (8.64 genes per Mb) compared with an aver-
age autosomal gene density of one gene per 103,622 bp
(9.65 genes per Mb). The average gene length of the
reference genome is 32,170 bp, while genes on X chro-
mosome scaffolds are on average 42,638 bp long (see
supplementary table 4,Supplementary Material online,
for further summary statistics). We used flow cytometry
data from Vanthournout et al. (2018) to estimate the
proportion of the genome made up by the X chromo-
somes. In S. africanus, we estimate that the X chromo-
somes make up 15.3% (SD: 0.009) of the total genome,
and in S. mimosarum it is 15.1% (SD: 0.012). We note that
the identified X chromosome scaffolds make up 9% of
the genome whereas flow cytometry indicates the X
chromosomes make up 15% of the genome.
Approximately half of the remaining 6% were not
assigned due to too low coverage from the RAD se-
quencing of the nuclei, while the other half are likely
to have a P
0
above the threshold of 0.238 (supplemen-
tary fig. 1,Supplementary Material online). We found no
differences in codon usage bias (Supek and Vlahovicek
2004) among genes located on X chromosomes and
autosomes in any of the two species (supplementary
fig. 2,Supplementary Material online).
Substitution Patterns of Autosomes
After processing the transcriptome sequence data, we
obtained consensus sequences of 4,641 putative orthologous
loci (including 523 on the X chromosomes) from all three
Stegodyphus species, and aligned these for comparative stud-
ies. We used PAML ver. 4.6 (Yang 2007)toestimatespecies-
specific dN/dSratios of S. mimosarum and S. africanus using S.
lineatus as outgroup for X chromosomes and autosomes sep-
arately. The autosomal dN/dSratio of the social S. mimosa-
rum was significantly larger than for S. africanus (0.131 vs.
0.114; randomization test: P¼0.004; fig. 3 and table 1), sug-
gesting stronger purifying selection in the outcrossing S. afri-
canus compared with the inbreeding S. mimosarum.Thisis
consistent with the estimate of a 10-fold higher effective pop-
ulationsizeintheoutcrossingS. africanus than the inbreeding
S. mimosarum (Settepani et al. 2017), and stronger effect of
selection in populations with larger effective size to remove
slightly deleterious mutations (Charlesworth 2009). A list of
genes assigned to X chromosome and autosome scaffolds can
be found in supplementary table 5,Supplementary Material
online.
Substitution Patterns of X Chromosome versus
Autosome
The X-linked dN/dSratios were 0.140 in S. africanus and 0.177
in S. mimosarum, and significantly larger than the autosomal
dN/dSratios of 0.114 in S. africanus and 0.131 in S. mimosa-
rum (randomization tests: P¼0.018 and P¼0.004, fig. 3 and
table 1). In both species, we found a significantly lower syn-
onymous substitution rate on the X-linked genes (dS
X
)com-
paredwithautosomalgenes(dS
A
)(forS. africanus dS
X
/dS
A
¼
0.72; randomization test: P<0.001; for S. mimosarum dS
X
/
dS
A
¼0.63; randomization test: P<0.001, table 1).
Genetic Diversity on X Chromosomes and Autosomes
From RAD sequencing, we obtained 24,321 RAD loci (3,440 X-
linked) from the outcrossing S. africanus and 20,665 RAD loci
(2,783 X-linked) from the inbreeding S. mimosarum.Using
the S. mimosarum reference genome (Sanggaard et al. 2014)
we found that 1.16% of the RAD loci are located in protein
coding regions. We estimated total diversity in three S. afri-
canus and five S. mimosarum populations, and found that all
S. mimosarum populations have reduced diversity on both X
chromosomes and autosomes compared with S. africanus
(both reduced by 85%) (fig. 4). We note that the diversity
estimates presented here are highly similar to those obtained
by Settepani et al. (2017) whoanalyzedthesameRADse-
quencedatausingadifferentpipeline.Variationindiversity
across scaffolds may reflect different rates of loss of diversity,
which is in accordance with linked selection playing a pre-
dominant role in loss of diversity. In S. mimosarum,onaver-
age 56% of the autosome scaffolds and 61% of the X-linked
scaffolds had a diversity of 0 (supplementary fig. 3,
Supplementary Material online), preventing us from mean-
ingful inference of the variation in loss of diversity across
Faster X Chromosome Evolution in Spiders .doi:10.1093/molbev/msz074 MBE
1283
Downloaded from https://academic.oup.com/mbe/article-abstract/36/6/1281/5420164 by Ghent University user on 11 March 2020
scaffolds, as the variation among these 0-diversity scaffolds is
lost due to the zero boundary. This was not the case for
S. africanus, where only 3% and 9% of autosome and X-linked
scaffolds, respectively, had diversity estimates of 0 (supple-
mentary fig. 3,Supplementary Material online), and we there-
fore contrasted variation in diversity across X chromosome
scaffolds to autosome scaffolds by the coefficient of variation
(CV). We find that diversity varies significantly more among
X-linked scaffolds than among autosome scaffolds consistent
with a stronger role for natural selection in removing diversity
in regions of the X chromosome than on the autosomes, by
either selective sweeps or background selection (fig. 5).
Estimates of piX/piA are more or less constant among sub-
social S. africanus populations, but varies significantly among
social S. mimosarum populations (S. africanus:F
2,99
¼0.003;
P¼0.99, S. mimosarum:F
4,146
¼4.63; P<0.01) (fig. 6).
Averaged across populations, the X to autosome diversity
ratio (piX/piA) is 0.57 (95% CI: 0.55–0.60) for S. africanus,
which is lower than the 0.75 expected with an equal contri-
bution of the two sexes. piX/piA of S. mimosarum,0.72(95%
CI: 0.65–0.79), was not significantly different from the 0.75
expected, but significantly higher than in S. africanus (v
2(1)
¼
4.25; P¼0.04) (fig. 6).
AA
AA
X1
Flowcytometry
‘Sample 0’‘Sample X1X2
AX A0
RAD sequencing
RAD sequences mapped to scaffolds
‘Sample 0’‘Sample X1X2
0( )=
#0( )
#0( )+# X1X2( )
Scaffold 1
Scaffold 2
Scaffold 3
Scaffold 4
Scaffold .
Scaffold .
Scaffold .
Scaffold n
Scaffold 1
Scaffold 2
Scaffold 3
Scaffold 4
Scaffold .
Scaffold .
Scaffold .
Scaffold n
0.0 0.2 0.4 0.6 0.8
0
Density
X2X1X2
FIG.2.Schematic presentation of study design including assignment of scaffolds to X or autosomes. Stegodyphus species, like most spiders, have an
X0 sex determination system, where males have only one copy of the sex chromosomes. Sperm cells were sorted into two pools using flow
cytometry: one with the sex chromosomes (“Sample X
1
X
2
”) and one without the sex chromosomes (“Sample 0”), and RAD sequencing libraries
from each pool were subsequently constructed and sequenced. The resulting RAD sequences from each pool were mapped to the scaffold
sequences of the S. mimosarum genome (Sanggaard et al. 2014). Scaffolds comprising the sex chromosomes were determined as the scaffolds with
no sequences (or few) mapping from the “Sample 0” pool, but with sequences mapping from the “Sample X
1
X
2
” pool. For each scaffold, we
estimated a summary statistic (P
0
) defined as the number of reads that mapped from “Sample 0” divided by the sum of reads that mapped from
both “Sample 0” and “Sample X
1
X
2
” after normalization of the total number of reads from both samples. Scaffolds belonging to X chromosomes are
predicted to have P
0
close to 0, while those belonging to autosomes are predicted to have P
0
close to 0.5.
dN/dS
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
XAXA
S. africanus
S. mimosarum
***
**
*
FIG.3.dN/dSestimates for X chromosomes and autosomes sepa-
rately from Stegodyphus africanus and S. mimosarum based on con-
sensus sequences of transcriptome data. Error bars represent 95%
confidence limits obtained by bootstrapping. Pvalues were estimated
by randomization tests. *<0.05, **<0.01.
Bechsgaard et al. .doi:10.1093/molbev/msz074 MBE
1284
Downloaded from https://academic.oup.com/mbe/article-abstract/36/6/1281/5420164 by Ghent University user on 11 March 2020
Population size fluctuations reduces effective population
size on X chromosomes relatively more than on autosomes
(Pool and Nielsen 2007,2008;Schou et al. 2017). Populations
of social spiders such as S. mimosarum undergo recurrent
population size fluctuations due to propagule dispersal of
single-mated females and high-population turnover rates
(Bilde et al. 2007;Settepani et al. 2014, 2017). Population
size fluctuations can therefore potentially explain the fluctu-
ating piX/piA among S. mimosarum populations. We used
simulations to quantitatively investigate the effects of recur-
rent population size fluctuations on piX/piA using fastsim-
coal2 (Excoffier et al. 2013). We found that recurrent
population size fluctuations with realistic parameters for S.
mimosarum can significantly reduce piX/piA (supplementary
fig. 4,Supplementary Material online). piX/piA is influenced
more by few founders (5 chromosomes vs. 50 chromosomes)
and to a lesser extent by slower growth after a founder event
(100 generations vs. 50 generations between founder events).
A McDonald–Kreitman test was performed using poly-
morphisms from the S. mimosarum RAD loci located in pro-
tein coding genes and divergences from the transcriptome
sequences. In total, we identified 150 synonymous and 140
nonsynonymous polymorphisms at the X chromosomes and
1,433 synonymous and 1,198 nonsynonymous polymor-
phisms at the autosomes. In the S. mimosarum lineage, we
identified 563 synonymous and 337 nonsynonymous substi-
tutions at the X chromosomes and 15,599 synonymous and
7,370 nonsynonymous substitutions at the autosomes. Alpha
estimates were estimated to be negative for both X chromo-
somes and autosomes (0.56 and 0.77, respectively), sug-
gesting that many nonsynonymous mutations are
segregating likely because they are slightly deleterious.
X Chromosome Substitution Rates and Sex-Biased
Mutation Rates
We find that the synonymous divergence of X chromosomes
is lower than for autosomes in both species estimated from
transcriptome data (dS
X
/dS
A
0.72 in S. africanus and 0.63 in S.
mimosarum;table 1). This divergence ratio does not account
for differences in coalescence times of X chromosomes and
autosomes caused by differences in effective population size
in the ancestor of S. africanus and S. mimosarum.Sincethe
effective population size of X chromosomes is smaller than
Table 1. dNand dSEstimates for Loci Located on the X Chromosomes and Autosome Scaffolds for Stegodyphus africanus and S. mimosarum.
dN(CI
95low
2CI
95high
)dS(CI
95low
2CI
95high
)dN/dS(CI
95low
2CI
95high
)
S. mimosarum
Autosomes 0.0012 (0.0012–0.0013) 0.0093 (0.0090–0.0096) 0.131 (0.125–0.137)
X chromosomes 0.0010 (0.0009–0.0012) 0.0059 (0.0052–0.0065) 0.177 (0.152–0.208)
S. africanus
Autosomes 0.0010 (0.0009–0.0010) 0.0083 (0.0080–0.0086) 0.114 (0.108–0.121)
X chromosomes 0.0008 (0.0007–0.0010) 0.0060 (0.0054–0.0066) 0.140 (0.120–0.164)
NOTE.—Stegodyphus lineatus was used as outgroup. In parenthesis are 95% confidence limits that are obtained by bootstrapping.
Genetic diversity (pi)
0.0000
0.0003
0.0006
0.0009
0.0012
0.0015
0.0018
0.0021
0.0024
0.0027
0.0030
0.0033
KRU
PON
WRF
MAH
PON
SAK
TANA
WEE
S. africanus (X)
S. africanus (A)
S. mimosarum (X)
S. mimosarum (A)
FIG.4. Genetic diversity estimated from RAD data for X chromosomes
(X) and autosomes (A) from three populations of Stegodyphus afri-
canus (WRF, PON and KRU) and five populations of S. mimosarum
(MAH, SAK, TANA, WEE and PON). We used data from between 5
and 10 individuals per loci from each population depending on cov-
erage. Error bars represent 95% confidence limits obtained by
bootstrapping.
Coefficient of variation of genetic diversity
S. africanus
0.0
0.4
0.8
1.2
All
WRF
PON
KRU
All
WRF
PON
KRU
A
X
FIG.5.Comparison of the coefficient of variation (CV) of genetic
diversity among scaffolds assigned to the X chromosomes and auto-
somes. Only in Stegodyphus africanus, was genetic diversity suffi-
ciently large to allow this comparison. WRF, PON and KRU
represent the three sampled S. africanus populations, while ALL is
the average per species. CVs were estimated as SD/average. Error bars
represent 95% confidence limits obtained by bootstrapping.
Faster X Chromosome Evolution in Spiders .doi:10.1093/molbev/msz074 MBE
1285
Downloaded from https://academic.oup.com/mbe/article-abstract/36/6/1281/5420164 by Ghent University user on 11 March 2020
that of autosomes, the X chromosomes are expected to co-
alesce faster than autosomes in the ancestral species. The
difference in synonymous divergence estimates of X chromo-
somesandautosomesisthereforenotsolelyduetodifferent
mutation rates, but also different times to accumulate sub-
stitutions. To correct for different coalescence times in the
ancestral species, we assumed an ancestral population size
(N
A
) of 300,000 (Settepani et al. 2017), a sex ratio of 1:1, and
species split time of 1 My (using the mutation rate from
Mattila et al. 2012). Under these assumptions, the predicted
time to coalescence of X chromosomes is 85% of that of the
autosomes (see supplementary fig. 5,Supplementary Material
online). Based on the adjusted dS
X
/dS
A
divergence ratio from
transcriptomedata(S. africanus: 0.85, S. mimosarum:0.74),we
estimate the male-to-female substitution ratio (a)(Miyata
et al. 1987)to2.6inS. africanus and 8.1 in S. mimosarum.
In addition, we calculated a synonymous divergence ratio
based on RAD data (dRAD
X
/dRAD
A
)inS. mimosarum,taking
advantage of the fact that the Madagascan and South African
populations are genetically isolated from each other. The es-
timated dRAD
X
/dRAD
A
divergence ratio is 0.85, and 0.89
when adjusting for different coalescence times in the ances-
tral population (supplementary fig. 5,Supplementary
Material online). Using the adjusted dRAD
X
/dRAD
A
diver-
gence ratio, we get an aestimate of 1.98.
Discussion
The method used to identify X-linked scaffolds in this
study is applicable for species with X0 or heterogametic
sex determination, where X chromosomes are
sufficiently large for sperm cells with and without the
X chromosomes to be separated using flow cytometry.
Large full-genome sequencing initiatives to sequence
5,000 insect and insect-related genomes (i5K) (Evans
et al. 2013), and the Global Invertebrate Genomics
Alliance (GIGA) (Bracken-Grissom et al. 2014)candi-
rectly benefit from our approach and allow a large num-
ber of sex chromosome systems to be investigated in
order to disentangle hypotheses regarding their involve-
ment in meiotic drive (Jaenike 2001;Unckless et al.
2015), sexual conflict (Andres and Morrow 2003;Mank
et al. 2014), and speciation (Presgraves 2008;Kitano et al.
2009).
Faster-X Evolution in S. mimosarum and S. africanus
We found evidence for faster-X evolution in both S. mimosa-
rum and S. africanus, providing the first case of faster-X evo-
lution in spiders (see supplementary table 1,Supplementary
Material online, for a survey of previous faster-X investiga-
tions) (Garrigan et al. 2014;Kousathanas et al. 2014;Sackton
et al. 2014). Faster-X can be caused by drift or adaptive sub-
stitutions at the X chromosomes. To test if faster-X is caused
by adaptive evolution, we used transcriptome data and RAD
sequences located in exons to estimate the proportion of
substitutions that are fixed by adaptive evolution using the
McDonald–Kreitman test (McDonald and Kreitman 1991).
Negative avalues were obtained for both X chromosomes
and autosomes, suggesting that slightly deleterious mutations
segregate. We can therefore not conclude from this analysis
to which extent faster-X is caused by drift or adaptive evolu-
tion. Estimating the proportion of adaptive substitution in
the presence of segregating slightly deleterious mutations
would require targeted sequencing of protein coding loci in
multiple individuals (Eyre-Walker and Keightley 2009). Two
other observations from our data are however informative
and consistent with adaptive evolution contributing to faster-
Xinthissystem.TheeffectivepopulationsizeofS. mimosa-
rum was reduced by 90% during the evolution of social
behavior (Settepani et al. 2017). Such an increase in genetic
drift has caused an increase in autosomal dN/dSof only 15%
(0.131 vs. 0.114). In comparison, a much lower difference in
effective population size of X chromosomes and autosomes is
associated with substantial increase in dN/dSof 35% (0.177 vs.
0.131) in S. mimosarum and 22% (0.140 vs. 0.114) in S. afri-
canus, supporting that the increase in dN/dSof X chromo-
somes is not only caused by genetic drift. Adaptive evolution
is further supported by the finding that diversity along the X
chromosomes varies more than along the autosomes in S.
africanus, suggesting that selective sweeps are more promi-
nent on the X chromosomes, a phenomenon also observed in
primates (Nam et al. 2017). Finally, in support of a prominent
role of drift causing faster-X, we find Ne
X
/Ne
A
<0.75 in S.
africanus (as estimated by piX/piA). However, if the difference
in diversity on X (piX) and A (piA) is caused by a lower
mutation rate at the X chromosomes and not drift, this is
unlikely to have an effect on adaptive substitutions (Vicoso
and Charlesworth 2009). Indeed our data suggests a lower X
chromosome mutation rate (see X Chromosome Mutation
piX/piA
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
All
KRU
PON
WRF
All
MAH
PON
SAK
TANA
WEE
S. africanus
S. mimosarum
a,b
a
a
b
a
FIG.6.Ratios of X chromosome to autosome genetic diversity (piX/
piA) based on pi estimates presented in figure 5. Estimates are pre-
sented for three populations of Stegodyphus africanus (WRF, PON
and KRU) and five populations of S. mimosarum (MAH, SAK, TANA,
WEE and PON), as well as the species average (ALL). Significant differ-
ences (indicated by different letters) between populations within
each species were investigated using F-tests and Tukey’s HSD method
for post hoc comparisons. Error bars represent 95% confidence limits
obtained by bootstrapping.
Bechsgaard et al. .doi:10.1093/molbev/msz074 MBE
1286
Downloaded from https://academic.oup.com/mbe/article-abstract/36/6/1281/5420164 by Ghent University user on 11 March 2020
Rate below), and the effects of genetic drift on dN/dSmay not
be as strong as suggested by the deviation of Ne
X
/Ne
A
from 0.75.
An alternative and nonexclusive explanation of faster-X is a
lower recombination rate of X chromosomes compared with
autosomes, arising as X chromosomes unlike autosomes only
recombine in females. A reduced recombination rate on X
chromosomes is predicted to increase the effect of linked
selection, which would increase dN/dSdue to fixation of
slightly deleterious mutations. This should produce a negative
correlation between recombination rate and rate of nonsy-
nonymous substitutions, as reported in, for example,
Drosophila (Assis et al. 2012).
The potential for “faster-X” evolution depends on the dif-
ference between the effective population sizes of X chromo-
somes (Ne
X
)andautosomes(Ne
A
). In species with a female-
biased sex ratio as observed in S. mimosarum, the difference
between Ne
X
and Ne
A
is expected to be lower compared with
species with equal sex ratio. Such a scenario provides a wider
range of dominance levels where beneficial mutations at the
X chromosomes are more rapidly fixed (Vicoso and
Charlesworth 2009), making species with biased sex ratio
more prone to “faster-X” evolution. However, according to
our diversity estimates of X chromosomes and autosomes, we
do not find support for intensified faster-X in S. mimosarum
relative to S. africanus (S. mimosarum,X
(dN/dS)
/A
(dN/dS)
: 1.35,
S. africanus,X
(dN/dS)
/A
(dN/dS)
: 1.22; P¼0.92). As S. mimosarum
was used for assignment of scaffolds to X chromosomes or
autosomes, usage of the same assignment in S. africanus and
thereby the species comparisons made above, relies on no
independent rearrangements occurring between X chromo-
somes and autosomes. Cytogenetic analyses of S. mimosarum
and S. africanus have shown that the X chromosomes appear
highly similar (Forman M, personal communication), sup-
porting the assumption that no major X chromosome rear-
rangements occurred since the species split.
Genetic Diversity on X Chromosomes and Autosomes
Our previous studies showed that the social species has a
much smaller effective population size and a high rate of
population turnover (Settepani et al. 2014, 2017). In agree-
ment with this, we observed considerably higher genetic di-
versity in S. africanus along with a lower dN/dSratio
suggesting that purifying selection is more efficient in the
outbreeding species.
The finding of an X to autosome diversity ratio (piX/piA) in
S. africanus lower than the expectation of 0.75 (no sex ratio
bias) suggests that additional evolutionary forces, such as
differences in mutation rates and/or selection may reduce
diversity on X chromosomes at a higher rate than on auto-
somes. Mutation rate on the X chromosomes was inferred to
be lower than on the autosomes, which at least partly explains
the low piX/piA in S. africanus. Selection is known to cause
loss of genetic diversity not only in the selected loci but also in
flanking regions due to genetic hitchhiking (Smith and Haigh
1974;Begun and Aquadro 1992) and background selection
(Charlesworth 2012). The effect of removing diversity by
linked selection is predicted to be larger in genomic regions
where recombination rates are small, as for X chromosomes
that do not recombine in males. The finding of lower diversity
on X chromosomes may therefore partly be due to selection.
Exposure of recessive variants on X chromosomes to selection
in males may enforce this effect, however, the lower effective
population size of the X chromosomes may cause selection to
be less efficient on X chromosome loci, potentially reducing
this effect. The social S. mimosarum has a primary female-
biased sex ratio (Lubin and Bilde 2007;Vanthournout et al.
2018), so a higher piX/piA is expected compared with the
subsocial S. africanus if the operational sex ratio is also female
biased (Ellegren 2009). In agreement with this expectation,
piX/piA in S. mimosarum was higher than in S. africanus.
Weproposethatthisisduetosimilarevolutionaryforcesas
discussed for S. africanus, which decrease piX/piA, and the
additional effect of female bias that increases piX/piA.
Social spiders are cooperative breeders with reproductive
skew so only a fraction of females reproduce (Lubin and Bilde
2007;Junghanns et al. 2017), but it is currently unclear how
large a proportion of females that reproduce, and therefore
what the operational sex ratio is. With everything else equal,
using the difference in piX/piA between the two species
makes it possible to estimate the operational sex ratio. The
point estimate of piX/piA (0.72) is consistent with an oper-
ational female bias between 1:8 and 1:9, and the lower bound-
ary of the confidence limits suggests that the operational
female bias is stronger than 1:2 (fig. 7). However, previous
studies suggest that the population sizes of social species
fluctuate substantially due to recurrent founder events asso-
ciated with population extinction/recolonization dynamics
(Crouch and Lubin 2001;Bilde et al. 2007;Settepani et al.
0.5
0.6
0.7
0.8
0.9
piX/piA standadized
1:1 1:2 1:3 1:4 1:5 1:6 1:7 1:8 1:9 1:10
Male to female sex ratio
S. mimosarum
S. africanus
Theoretical prediction
FIG.7. Theoretically expected piX/piA as a function of male to female
sex ratio (black dots). The expected piX/piA is standardized according
to the finding that Stegodyphus africanus has a lower piX/piA than the
expected 0.75, which we infer to be caused by a lower X chromosome
mutation rate and the effect of linked selection, and assume to have a
similar effect in S. mimosarum. Point estimates of piX/piA for the two
species are depicted with dotted lines, and for S. mimosarum, the 95%
confidence limits of this estimate are shown.
Faster X Chromosome Evolution in Spiders .doi:10.1093/molbev/msz074 MBE
1287
Downloaded from https://academic.oup.com/mbe/article-abstract/36/6/1281/5420164 by Ghent University user on 11 March 2020
2017). Population size fluctuations affect diversity on X chro-
mosomes more than diversity on autosomes, and therefore
also the piX/piA ratio due to the Pool–Nielsen effect (Pool
and Nielsen 2007). We simulated recurrent founder events
and showed that piX/piA values constantly lower than equi-
librium (estimated by simulating a constant population size)
can be reached when founder events are frequent. Depending
on the stage in the Pool–Nielsen cycle following a founder
event at which the dynamic equilibrium is modeled, popula-
tion dynamics with recurrent founder events would explain
the variation observed in piX/piA among the S. mimosarum
populations. The actual effect of female bias on piX/piA may
therefore be larger than we observed due to a possible coun-
teracting effect of population size fluctuations, and conse-
quently the operational sex ratio even more female biased
(supplementary fig. 6,Supplementary Material online).
X Chromosome Mutation Rate
We found lower synonymous divergence on X chromosomes
compared with autosomes suggesting a lower mutation rate
on X than for the autosomes in both species. Importantly, this
finding persisted when taking differences in coalescence time
of X chromosomes and autosomes in the ancestral species
into account. A lower mutation rate on X chromosomes can
have several causes. One possibility is that the mutation rate is
lower on the X chromosomes simply due to different se-
quence composition of the X chromosomes and autosomes.
However, to our knowledge there is no evidence from previ-
ous studies that this is a plausible explanation. Another pos-
sible cause is a lower recombination rate of X chromosomes,
which only recombine in females, than of autosomes which
recombine in both sexes. Since recombination can be a source
of mutations (Arbeithuber et al. 2015), the mutation rate on
X chromosomes is expected to be lower than on autosomes,
but the overall effect is not expected to be very high due to
the relatively low number of recombination events on the X-
chromosome per generation. Another possibility is that
mutations are male-biased. Since X chromosomes spend 1/
3 of their time in males and 2/3 in females, a male-biased
mutation rate will cause a lower mutation rate at X chromo-
some compared with autosomes. A male-biased mutation
rate has been found in several vertebrate species, and has
been interpreted to be due to more cell divisions in spermato-
genesis than in oogenesis. However, in short lived species like
spiders and other invertebrates, the number of cell divisions
in spermatogenesis and oogenesis is often similar, like in, for
example, Drosophila (Drost and Lee 1998). Recent evidence
mainly from humans suggests an alternative cause of a male-
biased mutation rate, namely that mutation rate simply is
higher in spermatogenesis than in oogenesis. No evidence for
such an effect in spiders exists, but as we find none of the
aforementioned explanations to be convincing, we suggest
this alternative as a possibility.
Conclusions
This first analysis of DNA sequence evolution of X chro-
mosomes in spiders reveals faster-X evolution in two
sister-species that differ in mating systems, population
dynamics and sex ratio bias. The extent of faster-X evo-
lution is similar in the two species, contrary to theoret-
ical predictions when sex ratios diverge from 50% to 50%.
Contrasting the relative genetic diversity on X chromo-
somes and autosomes in a social inbreeding and a sub-
social outcrossing species revealed higher piX/piA, and
larger variation of piX/piA among populations in the
social inbreeding species. These findings are consistent
with the effects of female bias and Pool–Nielsen effects
caused by frequent population size fluctuations in the
social inbreeding S. mimosarum. Finally, we infer that the
X chromosome mutation rate is lower than the auto-
some mutation rate in both species, potentially caused
by a higher mutation rate in spermatogenesis than in
oogenesis.
Materials and Methods
Study System
The spider genus Stegodyphus (family Eresidae) contains
more than 20 species. Three of the species have an indepen-
dently derived social behavior (fig. 1)(Johannesen et al. 2007;
Settepani et al. 2016), which is consistently associated with a
female-biased sex ratio, reproductive skew and an inbreeding
mating system, also named the “social syndrome” (Lubin and
Bilde 2007). Family groups of social species live and breed in
closed nests that propagate within populations by nest fission
and by long distance dispersal through ballooning of mated
females (Lubin and Bilde 2007). In comparison, the subsocial
species have equal sex ratios, no reproductive skew and are
outcrossing (Bilde et al. 2005;Lubin and Bilde 2007).
Data Sets
RAD Sequence Data of Sperm Cells
To allocate reference scaffolds to an autosome or an X chro-
mosome, we used flow cytometry (Garner et al. 2013)tosort
free nuclei from S. mimosarum sperm cells, and subsequently
RAD sequencing the DNA (fig. 2). Free nuclei from sperm cells
were obtained by trypsin treatment and their DNA was
stainedwithpropidiumiodide(Vindelov et al. 1983;Aron
et al. 2003;Vanthournout et al. 2014).Thenucleiweresorted
based on DNA content into on a BD Biosciences FACSAria
cell sorter (Argon laser emitting at 488 nm), into a sample
with the two X chromosomes (“Sample X
1
X
2
”) and one with-
out the two X chromosomes (“Sample 0”). From each of the
two samples, paired-end RAD sequencing libraries were con-
structed using the protocol described in (Poland et al. 2012),
with the following modifications: 0.5 mlofBSAwasaddedto
the Restriction Mastermix and an AMPure Beads clean-up
and size selection step was implemented after PCR amplifi-
cation. The libraries were sequenced using the Illumina HiSeq
2000 platform (100 bp paired-end).
Transcriptome Sequence Data
To enable inference of the coding substitution patterns be-
tween S. africanus and S. mimosarum, we obtained transcrip-
tomes of the two species as well as an outgroup species (S.
lineatus). Libraries of an S. lineatus and an S. africanus female
Bechsgaard et al. .doi:10.1093/molbev/msz074 MBE
1288
Downloaded from https://academic.oup.com/mbe/article-abstract/36/6/1281/5420164 by Ghent University user on 11 March 2020
were constructed using Illumina’s TruSeq Stranded mRNA LT
Sample Prep Kit, and sequenced on an Illumina HiSeq2000
platform (100 bp paired-end). For S. mimosarum, quality fil-
tered transcriptome data, also sequenced on an Illumina
HiSeq 2000 platform (100 bp paired-end), from a previous
study was used (Sanggaard et al. 2014).
RAD Sequence Data of Populations
To estimate molecular diversity of autosomes and X chromo-
somes in S. mimosarum and S. africanus, we used quality
filtered RAD sequenced reads (100 bp paired-end) from a
previously published study (Settepani et al. 2017). This data
set contained individual data from 49 S. mimosarum females
(each sampled from its own distinct nest) and 27 S. africanus
females, with an average of 3.6 million clean reads per indi-
vidual. The S. mimosarum females were sampled from five
populations (fig. 1); ten from each of four populations (MAH,
SAK, TANA, WEE) and nine from one population (PON).
Three of the S. mimosarum populations are located in
Madagascar (MAH, SAK, TANA) and two in South Africa
(WEE, PON). The S. africanus females were sampled from
three South African populations (WRF, PON, KRU) (fig. 1),
with eight, ten, and nine, respectively).
Data Analyses
Identifying Scaffolds from the X Chromosomes
The RAD sequence reads from the sorted sperm cells were
quality trimmed using the FASTX toolkit (http://hannonlab.
cshl.edu/fastx_toolkit). We discarded reads containing a base
with a Phred quality score <10,aswellasreadswithan
average Phred quality score <30. More clean reads were
obtained from “Sample 0” than “Sample X
1
X
2
,” and it was
therefore subsampled to obtain same number of reads from
both samples (1,034,261 reads). The clean data from the two
samples (“Sample 0” and “Sample X
1
X
2
”) were mapped sep-
arately to the reference genome sequence of S. mimosarum
(Sanggaard et al. 2014), using CLC Genomics Workbench 7
(default parameters). The reference genome consists of
23,000 scaffolds (N50 ¼480,636 bp) and 45,000 contigs
(N50 ¼17,272 bp). Scaffolds for which at least 100 reads
from “Sample X
1
X
2
” mapped (3,490 in total), was considered
tohaveapotentialXchromosomeorigin.Thenumberof
reads mapped from “Sample 0” was hereafter divided by the
sum of the corrected number of reads mapped from “Sample
0” and “Sample X
1
X
2
.” We call this proportion P
0
.
P0iðÞ¼ #reads Sample 0ðiÞ
#reads Sample 0ðiÞþ#reads Sample X1X2ðiÞ:
Given a high-quality sorting of nuclei, we expect a bimodal
distribution of P
0
. One mode will contain the distribution for
the scaffolds belonging to the X chromosomes and the other
will contain the distribution for the scaffolds belonging to the
autosomes. The first distribution is expected to have an av-
erage just >0, as the sorting of the two nuclei types is imper-
fect. The second distribution is expected to have an average
just >0.5, since the reads from “Sample X
1
X
2
” will be mapped
to more scaffolds (both X chromosomes and autosomes
scaffolds) than the reads from “Sample 0” (only autosomes).
How much >0.5 depends on the distribution of RAD loci on
X chromosomes and autosomes, and the precision of the
sorting of the two nuclei types. A Bayesian mixture analysis
was performed in order to separate the two distributions and
estimate their proportion, mean, and variance (fig. 2; see also
Supplementary Material online). The two identified distribu-
tions overlapped slightly, with 0.119 and 0.500 as the means of
the distribution predicted to be composed of X chromosome
and autosome scaffolds, respectively. When determining
what minimum P
0
to use as a cut-off for a scaffold to be
considered an autosome, we tested three different cut-offs
(0.119, 0.3, and 0.5). This was necessary due to the overlap of
distributions in the autosomal peak (supplementary fig. 1,
Supplementary Material online). Using transcriptomic data
and RAD sequencing data on each of the three cut-offs, we
estimated pi and dN/dSratio for the autosomes, respectively.
These two measures were consistent across the three cut-offs,
and in particular between 0.3 and 0.5 (supplementary figs. 7
and 8,Supplementary Material online). Based on this com-
parison, and an estimated false positive rate of 2.5% (supple-
mentary table 2,Supplementary Material online) we
continued with P
0
¼0.3 as the lower cut-off for the full set
of analyses. A false positive rate of 2.5% was also used as to
determine the upper cut-off for assigning scaffolds as belong-
ing to the X chromosomes (P
0
¼0.239).
Molecular Evolution at X Chromosomes and Autosomes
The raw sequences of all three species were quality trimmed
using the FASTX toolkit (http://hannonlab.cshl.edu/fastx_
toolkit). We discarded reads containing a position with a
Phred quality score <10, as well as reads with an average
Phred quality score <30. The clean data from all three species
were mapped separately to a gene list of the S. mimosarum
genome (Sanggaard et al. 2014) consisting of 26,314 loci all
beginning with a start codon (ATG) using CLC Genomics
Workbench 7 (default parameters). For each species, loci
with average coverage <3wereinitiallyremoved.
Consensus bases were called in all positions with coverage
8 or higher, while positions with coverage between 3 and 7
were masked and not included in downstream analyses.
Ambiguous bases (IUPAC) were called when a base was sup-
ported by at least three reads and/or if its proportion was
>10%. Only consensus sequences with <2.5% ambiguous
bases were retained. The resulting consensus sequences
were grouped based on their mapping to the S. mimosarum
gene list by assuming orthology and subsequently aligned
across species using PRANK (Loytynoja and Goldman
2008). Alignments were manually edited assuming that frame
shifts were caused by sequencing or assembly errors. In total,
8,302 alignments with sequences from all three species were
obtained, of which 285 belonged to X chromosome scaffolds
and 8,017 to autosome scaffolds. All codons that could not be
translated into an amino acid for a given species (because of
Ns or ambiguous nucleotides) were identified, and the codons
were removed from all three species. Synonymous (dS)sub-
stitution rates, nonsynonymous (dN) substitution rates, and
Faster X Chromosome Evolution in Spiders .doi:10.1093/molbev/msz074 MBE
1289
Downloaded from https://academic.oup.com/mbe/article-abstract/36/6/1281/5420164 by Ghent University user on 11 March 2020
dN/dSratios were estimated for X chromosomes and auto-
somes separately in both S. mimosarum and S. africanus using
PAML ver. 4.6 (Yang 2007). 95% confidence limits of dN,dS,
and dN/dSwere estimated by bootstrapping over the genes
(n¼1,000) and producing one overall (across genes) estimate
of dN,dS,anddN/dSfor each sampling.
Molecular Diversity at X Chromosomes and Autosomes
A RAD reference was constructed for both S. mimosarum and
S. africanus in two steps. First, all sequences represented by at
least three identical reads were obtained using a custom pro-
gram in all individuals separately (“clc_find_maximal,” see
Supplementary Material online for more information). In
the second step, species-specific RAD references were created
by grouping all the resulting sequences from conspecific indi-
viduals with >98% similarity using a custom program
(“clc_find_groups,” for more detail, see Supplementary
Material online). The resulting RAD reference sets were
mapped to the genome sequence of S. mimosarum
(Sanggaard et al. 2014), and in cases with more than one
reference sequence mapped to same position, all but one
were removed using a custom script (“remove_dup.tcsh,”
for more detail, see Supplementary Material online), giving
two final RAD reference sets, one for each species. Each indi-
vidual was subsequently mapped to these RAD reference sets
using CLC Genomics Workbench 7, and consensus sequences
were extracted from all mappings having a minimum of
8 reads and a maximum of 40 reads. Ambiguous bases
(IUPAC) were called when the least frequent base was sup-
ported by at least three reads and/or if its proportion was
>10%, but also considering the read quality score as imple-
mented in the CLC Genomics Workbench 7. If a consensus
sequence had >2.5% ambiguous bases it was discarded as-
suming that the mapped reads originated from more than
one genomic position. For each consensus sequence, we called
two alleles that were subsequently aligned per locus for all
individuals within each population using PRANK (Loytynoja
and Goldman 2008). We discarded alignments with less than
five individuals represented, and separated remaining align-
ments into X chromosome and autosome based on the map-
ping of the RAD reference sets to the S. mimosarum genome
scaffolds (see above). Since S. africanus is closely related to S.
mimosarum, we interpreted the S. africanus RADs mapping to
S. mimosarum XchromosomescaffoldsasXchromosomesin
S. africanus as well. For each set of alignments, all the align-
ments were concatenated, with missing sequences written as
gaps, and split into equally long subalignments. This length
was identical across all populations, X chromosome sets and
autosome sets. This universal length of the subalignments was
set such that the population with smallest X chromosome
coverage had 15 subalignments (KRU: 136,432 bp). Within
population genetic diversity (Tajima 1983) was calculated
for each subalignment using the package ape in R (Paradis
et al. 2004;R Core Team 2015), and average pi was calculated
for X chromosomes and autosomes for each population. 95%
confidence limits were estimated by bootstrapping over the
sub alignments (n¼10,000).
McDonald–Kreitman Test
Rad sequence data from all S. mimosarum individuals
were mapped to the reference genome using bwa (Li
and Durbin 2009). Polymorphic positions were called in
positions with minimum coverage of 10using Samtools
and bcftools (Li et al. 2009). Sites that were polymorphic
intheRADsequencedataandsitesthatdifferedbetween
RADsequencedataandthereferencegenomewerecon-
sidered. snpEff was used to identify variants located in
protein coding positions, and if they were synonymous
or nonsynonymous (Cingolani et al. 2012). The number of
synonymous and nonsynonymous substitutions on X
chromosomes and autosomes were taken from PAML
analyses described earlier.
Variation in Diversity between X Chromosomes and
Autosomes
Alignments of RAD loci from separate scaffolds were
concatenated, and pi was estimated per scaffold with three
or more RAD loci using the package ape (Paradis et al. 2004)
in R (R Core Team 2016). Coefficient of variation was esti-
mated for autosome and X chromosome scaffolds separately
for each population using the estimator SD/average.
Simulations
We simulated DNA sequences under a recurrent bottleneck
scenario using fastsimcoal2 (Excoffier et al. 2013). This was
done with two different population sizes, 20,000 and 15,000
representing autosomes and X chromosomes, respectively,
and using a mutation rate of 1.2E-8. These parameters were
chosen to reach diversity similar to the estimates obtained
from the RAD sequence data. Data were simulated to mimic
our RAD sequence data by simulating 20,000 independent
loci of 100 bp. Data were simulated under four different bot-
tlenecks scenarios; 50 and 100 generations between bottle-
necks combined with magnitude of bottlenecks of 1% and
10% (supplementary fig. 9,Supplementary Material online). Pi
was estimated at different time points using the package ape
(Paradis et al. 2004)inR(R Core Team 2016).
Statistical Analyses
When testing for species differences in piX/piA it was neces-
sary to account for populations using a random effect in a
mixed model in the package lme4 (Bates et al. 2015)inR(R
Core Team 2016). Data used for this test were the subalign-
ments of equal length created for the bootstrapping (see
above). For each population, we grouped the autosomal sub-
alignments in as many groups as there were subalignments of
the X chromosomes. We then calculated the median (due to
a highly skewed distribution of pi in S. mimosarum which has
high frequency of scaffolds with zero pi) pi for each autosomal
group and paired it randomly with an X chromosome sub-
alignment, in this way, we got several independent estimates
of piX/piA for each population. The statistical significance of
the effect of species was assessed using a likelihood ratio test.
To further test for population differentiation within species,
we constructed a linear model for each species, containing
Bechsgaard et al. .doi:10.1093/molbev/msz074 MBE
1290
Downloaded from https://academic.oup.com/mbe/article-abstract/36/6/1281/5420164 by Ghent University user on 11 March 2020
population as the only predictor variable, and used F-tests
and Tukey’s HSD method for post hoc comparisons. To test
for differences in dN/dSbetween autosomes and X chromo-
somes within a species, we used a randomization test in
which we permuted chromosomal origin (autosome or X
chromosome) of all the genes (npermutations ¼1,000)
and estimated one overall (across genes) estimate of dN/dS
for genes assigned to autosome or X chromosome in each
sampling. We then estimated X
(dN/dS)
A
(dN/dS)
for each per-
mutation, and used this as a test-statistic in a two-tailed test
by comparison to the observed difference. We used the same
approach when testing for differences in dS, but here we used
the ratio dS
X
/dS
A
as a test-statistic. A similar approach was
used to test for differences in A
(dN/dS)
or in X
(dN/dS)
between
the two species, except here we permuted the species origin
of each gene and used the test-statistics A
africanus (dN/dS)
A
mimosarum (dN/dS)-
and X
africanus (dN/dS)
X
mimosarum (dN/dS)
.
Finally, when analyzing whether the difference in dN/dSbe-
tween X-linked and autosomal genes differs between S. mim-
osarum and S. africanus,weused(X
(dN/dS)
/A
(dN/dS)
)
africanus
(X
(dN/dS)
/A
(dN/dS)
)
mimosarum
as a test-statistic.
X Chromosome Mutation Rate. We estimated the male-to-
female mutation rate from synonymous divergence under
the assumption that the synonymous mutation rate equals
the synonymous substitution rate (Kimura 1983), by compar-
ing synonymous divergence at X chromosomes and auto-
somes. We used the formula k
x
/k
A
¼(2/3)(2þa)/(1þa)
following Miyata et al. (1987), where k is the synonymous
sequence divergence. We obtained two independent esti-
mates; 1) synonymous divergence (dS) between the two spe-
cies estimated from transcriptome data, and 2) divergence
between the two genetically isolated groups of S. mimosarum
populations from Madagascar and South Africa, respectively,
estimated from RAD data. The latter analysis was done under
the assumption that RAD loci evolve neutrally. To adjust for
different times to accumulate substitutions for X chromo-
somes and autosomes due to different coalescence times of
X chromosomes and autosomes in the ancestral species/pop-
ulation caused by differences in Ne, we used coalescence
calculations to adjust k
x
/k
A
. These calculations are based on
coalescence times in the ancestral species/population, which
is a function of effective ancestral population sizes (N
A
).
Approximate N
A
estimates were obtained from Settepani
et al. (2017).
Supplementary Material
Supplementary data areavailableatMolecular Biology and
Evolution online.
Acknowledgments
The study was supported by a grant from the European
Research Council (ERC StG-2011_282163) to T.B. We thank
the FACS Core Facility (Aarhus University) where flow cyto-
metric sorting was performed. We also thank Anne Aagaard
Lauridsen and Shenglin Liu for useful suggestions to the man-
uscript. We thank Marie Rosenstand Hansen for technical
assistance. Sequence data can be downloaded from NCBI
under the BioProject ID: PRJNA453114.
References
Andres JA, Morrow EH. 2003. The origin of interlocus sexual conflict: is
sex-linkage important? J Evol Biol. 16(2):219–223.
Arbeithuber B, Betancourt AJ, Ebner T, Tiemann-Boege I. 2015.
Crossovers are associated with mutation and biased gene conversion
at recombination hotspots. Proc Natl Acad Sci U S A.
112(7):2109–2114.
Arbiza L, Gottipati S, Siepel A, Keinan A. 2014. Contrasting X-linked and
autosomal diversity across 14 human populations. Am J Hum Genet.
94(6):827–844.
Aron S, De Menten L, Bockstaele DV. 2003. Brood sex ratio determina-
tion by flow cytometry in ants. MolEcolResourc. 3:471–475.
AssisR,ZhouQ,BachtrogD.2012.Sex-biased transcriptome evolution in
Drosophila. Genome Biol Evol. 4(11):1189–1200.
Bates D, Machler M, Bolker BM, Walker SC. 2015. Fitting linear mixed-
effectsmodelsusinglme4.JStatSoftw. 67:1–48.
Begun DJ, Aquadro CF. 1992. Levels of naturally occurring DNA poly-
morphism correlate with recombination rates in D. melanogaster.
Nature 356(6369):519–520.
Betancourt AJ, Presgraves DC, Swanson WJ. 2002. A test for faster X
evolutioninDrosophila.Mol Biol Evol. 19(10):1816–1819.
Bilde T, Coates KS, Birkhofer K, Bird T, Maklakov AA, Lubin Y, Aviles L.
2007. Survival benefits select for group living in a social spider despite
reproductive costs. JEvolBiol. 20(6):2412–2426.
Bilde T, Lubin Y, Smith D, Schneider JM, Maklakov AA. 2005. The tran-
sition to social inbred mating systems in spiders: role of inbreeding
tolerance in a subsocial predecessor. Evolution 59(1):160–174.
Bracken-Grissom H, Collins AG, Collins T, Crandall K, Distel D, Dunn C,
Giribet G, Haddock S, Knowlton N, Martindale M, et al. 2014. The
Global Invertebrate Genomics Alliance (GIGA): developing commu-
nity resources to study diverse invertebrate genomes. JHered.
105(1):1–18.
Carneiro M, Albert FW, Melo-Ferreira J, Galtier N, Gayral P, Blanco-
Aguiar JA, Villafuerte R, Nachman MW, Ferrand N. 2012. Evidence
for widespread positive and purifying selection across the European
rabbit (Oryctolagus cuniculus)genome.MolBiolEvol.
29(7):1837–1849.
Charlesworth B. 2009. Effective population size and patterns of molec-
ular evolution and variation. Nat Rev Genet. 10(3):195–205.
Charlesworth B. 2012. The effects of deleterious mutations on evolution
at linked sites. Genetics 190(1):5–22.
Charlesworth B, Campos JL, Jackson BC. 2018. Faster-X evolution: theory
and evidence from Drosophila. Mol Ecol. 27(19):3753–3771.
Charlesworth B, Coyne JA, Barton NH. 1987. The relative rates of evo-
lution of sex-chromosomes and autosomes. Am Nat.
130(1):113–146.
Cingolani P, Platts A, Wang LL, Coon M, Nguyen T, Wang L, Land SJ, Lu
XY, Ruden DM. 2012. A program for annotating and predicting the
effects of single nucleotide polymorphisms, SnpEff: SNPs in the ge-
nome of Drosophila melanogaster strain w(1118); iso-2; iso-3. Fly
6(2):80–92.
Counterman BA, Ortiz-Barrientos D, Noor M. 2004. Using comparative
genomic data to test for fast-x evolution. Evolution 58(3):656–660.
Crouch T, Lubin Y. 2001. Population stability and extinction in a social
spider Stegodyphus mimosarum (Araneae: eresidae). Biol J Linn Soc.
72(3):409–417.
Cutter AD, Wasmuth JD, Washington NL. 2008. Patterns of molecular
evolution in Caenorhabditis preclude ancient origins of selfing.
Genetics 178(4):2093–2104.
Drost JB, Lee WR. 1998. The developmental basis for germline mosaicism
in mouse and Drosophila melanogaster.Genetica 102-3:421–443.
Ellegren H. 2007. Characteristics, causes and evolutionary consequences
of male-biased mutation. Proc R Soc B. 274(1606):1–10.
Ellegren H. 2009. The different levels of genetic diversity in sex chromo-
somes and autosomes. Trends Genet. 25(6):278–284.
Faster X Chromosome Evolution in Spiders .doi:10.1093/molbev/msz074 MBE
1291
Downloaded from https://academic.oup.com/mbe/article-abstract/36/6/1281/5420164 by Ghent University user on 11 March 2020
Ellegren H, Galtier N. 2016. Determinants of genetic diversity. Nat Rev
Genet. 17(7):422–433.
Evans JD, Brown SJ, Hackett KJ, Robinson G, Richards S, Lawson D, Elsik C,
Coddington J, Edwards O, Emrich S. 2013. The i5K initiative: advanc-
ing arthropod genomics for knowledge, human health, agriculture,
and the environment i5K Consortium. J Hered. 104:595–600.
Excoffier L, Dupanloup I, Huerta-Sanchez E, Sousa VC, Foll M. 2013.
Robust demographic inference from genomic and SNP data. PLoS
Genet. 9(10):e1003905.
Eyre-Walker A, Keightley PD. 2009. Estimating the rate of
adaptive molecular evolution in the presence of slightly
deleterious mutations and population size change. Mol Biol Evol.
26(9):2097–2108.
Garner DL, Evans KM, Seidel GE. 2013. Sex-sorting sperm using flow
cytometry/cell sorting. Methods Mol Biol. 927:279–295.
Garrigan D, Kingan SB, Geneva AJ, Vedanayagam JP, Presgraves DC. 2014.
Genome diversity and divergence in Drosophila mauritiana:multiple
signatures of Faster X evolution. Genome Biol Evol. 6(9):2444–2458.
Guo YL, Bechsgaard JS, Slotte T, Neuffer B, Lascoux M, Weigel D,
Schierup MH. 2009. Recent speciation of Capsella rubella from
Capsella grandiflora, associated with loss of self-incompatibility
and an extreme bottleneck. Proc Natl Acad Sci U S A.
106(13):5246–5251.
Hedrick PW. 2007. Sex: differences in mutation, recombination, selection,
gene flow, and genetic drift. Evolution 61(12):2750–2771.
Hu TT, Eisen MB, Thornton KR, Andolfatto P. 2013. A second-generation
assembly of the Drosophila simulans genome provides new insights
into patterns of lineage-specific divergence. Genome Res.
23(1):89–98.
Hutter S, Li HP, Beisswanger S, De Lorenzo D, Stephan W. 2007. Distinctly
different sex ratios in African and European populations of
Drosophila melanogaster inferred from chromosomewide single nu-
cleotide polymorphism data. Genetics 177(1):469–480.
Hvilsom C, Qian Y, Bataillon T, Li YR, Mailund T, Salle B, Carlsen F, Li RQ,
Zheng HC, Jiang T, et al. 2012. Extensive X-linked adaptive evolution
in central chimpanzees. Proc Natl Acad Sci U S A. 109(6):2054–2059.
Jaenike J. 2001. Sex chromosome meiotic drive. Annu Rev Ecol Syst.
32(1):25–49.
Jaquiery J, Peccoud J, Ouisse T, Legeai F, Prunier-Leterme N, Gouin A,
Nouhaud P, Brisson JA, Bickel R, Purandare S, et al. 2018.
Disentangling the causes for Faster-X evolution in Aphids. Genome
Biol Evol. 10(2):507–520.
Jaquiery J, Stoeckel S, Rispe C, Mieuzet L, Legeai F, Simon JC. 2012.
Accelerated evolution of sex chromosomes in aphids, an X0 system.
MolBiolEvol. 29(2):837–847.
Johannesen J, Lubin Y, Smith DR, Bilde T, Schneider JM. 2007. The age
and evolution of sociality in Stegodyphus spiders: a molecular phy-
logenetic perspective. Proc R Soc B. 274(1607):231–237.
Junghanns A, Holm C, Schou MF, Sorensen AB, Uhl G, Bilde T. 2017.
Extreme allomaternal care and unequal task participation by
unmated females in a cooperatively breeding spider. Anim Behav.
132:101–107.
Kimura M. 1983. The neutral theory of molecular evolution. Cambridge:
Cambridge University Press.
KitanoJ,RossJA,MoriS,KumeM,JonesFC,ChanYF,AbsherDM,
Grimwood J, Schmutz J, Myers RM, et al. 2009. A role for a neo-sex
chromosome in stickleback speciation. Nature
461(7267):1079–1083.
Kousathanas A, Halligan DL, Keightley PD. 2014. Faster-X adaptive pro-
teinevolutioninhousemice.Genetics 196(4):1131.
Li H, Durbin R. 2009. Fast and accurate short read alignment with
Burrows-Wheeler transform. Bioinformatics 25(14):1754–1760.
LiH,HandsakerB,WysokerA,FennellT,RuanJ,HomerN,MarthG,
Abecasis G, Durbin R, Genome Project Data P. 2009. The Sequence
Alignment/Map format and SAMtools. Bioinformatics
25(16):2078–2079.
Loytynoja A, Goldman N. 2008. Phylogeny-aware gap placement pre-
vents errors in sequence alignment and evolutionary analysis. Science
320(5883):1632–1635.
Lu J, Wu CI. 2005. Weak selection revealed by the whole-genome com-
parison of the X chromosome and autosomes of human and chim-
panzee. Proc Natl Acad Sci U S A. 102(11):4063–4067.
Lubin Y, Bilde T. 2007. The evolution of sociality in spiders. Advances in
the study of behavior, 37: 83–145.
Mank JE, Axelsson E, Ellegren H. 2007. Fast-X on the Z: rapid evolution of
sex-linked genes in birds. Genome Res. 17(5):618–624.
Mank JE, Hosken DJ, Wedell N. 2014. Conflict on the sex chromosomes:
cause, effect, and complexity. Cold Spring Harb Perspect Biol.6.
Mank JE, Nam K, Ellegren H. 2010. Faster-Z evolution is predominantly
due to genetic drift. Mol Biol Evol. 27(3):661–670.
Mattila TM, Bechsgaard JS, Hansen TT, Schierup MH, Bilde T. 2012.
Orthologous genes identified by transcriptome sequencing in the
spider genus Stegodyphus.BMC Genomics 13:70.
McDonald JH, Kreitman M. 1991. Adaptive protein evolution ar the adh
locus in Drosophila. Nature 351(6328):652–654.
Miyata T, Hayashida H, Kuma K, Mitsuyasu K, Yasunaga T. 1987. Male-
driven molecular evolution – a model and nucleotide sequence
analysis. Cold Spring Harb Symp Quant Biol. 52:863–867.
Nam K, Munch K, Mailund T, Nater A, Greminger MP, Krutzen M,
Marques-Bonet T, Schierup MH. 2017. Evidence that the rate of
strong selective sweeps increases with population size in the great
apes. Proc Natl Acad Sci U S A. 114(7):1613–1618.
Paradis E, Claude J, Strimmer K. 2004. APE: analyses of phylogenetics and
evolution in R language. Bioinformatics 20(2):289–290.
Poland JA, Brown PJ, Sorrells ME, Jannink JL. 2012. Development of high-
density genetic maps for barley and wheat using a novel two-
enzyme genotyping-by-sequencing approach. PLoS One 7(2):e32253.
Pool JE, Corbett-Detig RB, Sugino RP, Stevens KA, Cardeno CM, Crepeau
MW, Duchen P, Emerson JJ, Saelao P, Begun DJ, et al. 2012.
Population genomics of sub-Saharan Drosophila melanogaster:
African diversity and non-African admixture. PLoS Genet.
8(12):e1003080.
Pool JE, Nielsen R. 2007. Population size changes reshape genomic pat-
terns of diversity. Evolution 61(12):3001–3006.
Pool JE, Nielsen R. 2008. The impact of founder events on chromosomal
variability in multiply mating species. Mol Biol Evol. 25(8):1728–1736.
Presgraves DC. 2008. Sex chromosomes and speciation in Drosophila.
Trends Genet. 24(7):336–343.
R Core Team. 2015. R: a language and environment for statistical com-
puting. Vienna (Austria): R Foundation for Statistical Computing.
R Core Team. 2016. R: a language and environment for statistical com-
puting. Vienna (Austria): R Foundation for Statistical Computing.
Rousselle M, Faivre N, Ballenghien M, Galtier N, Nabholz B. 2016.
Hemizygosity enhances purifying selection: lack of Fast-Z evolution
in two Satyrine butterflies. Genome Biol Evol. 8(10):3108–3119.
Sackton TB, Corbett-Detig RB, Nagaraju J, Vaishna L, Arunkumar KP,
Hartl DL. 2014. Positive selection drives Faster-Z evolution in silk-
moths. Evolution 68(8):2331–2342.
Salomon M, Lubin Y. 2007. Cooperative breeding increases reproductive
success in the social spider Stegodyphus dumicola (Araneae,
Eresidae). Behav Ecol Sociobiol. 61(11):1743–1750.
Sanggaard KW, Bechsgaard JS, Fang XD, Duan JJ, Dyrlund TF, Gupta V,
Jiang XT, Cheng L, Fan DD, Feng Y, et al. 2014. Spider genomes
provide insight into composition and evolution of venom and silk.
Nat Commun. 5:Article number: 3765.
Schou MF, Loeschcke V, Bechsgaard J, Schlotterer C, Kristensen TN. 2017.
Unexpected high genetic diversity in small populations suggests
maintenance by associative overdominance. Mol Ecol.
26(23):6510–6523.
Settepani V, Bechsgaard J, Bilde T. 2014. Low genetic diversity and strong
but shallow population differentiation suggests genetic homogeni-
zation by metapopulation dynamics in a social spider. JEvolBiol.
27(12):2850–2855.
Settepani V, Bechsgaard J, Bilde T. 2016. Phylogenetic analysis suggests
that sociality is associated with reduced effectiveness of selection.
Ecol Evol. 6(2):469–477.
Settepani V, Schou MF, Greve M, Grinsted L, Bechsgaard J, Bilde T.
2017. Evolution of sociality in spiders leads to depleted genomic
Bechsgaard et al. .doi:10.1093/molbev/msz074 MBE
1292
Downloaded from https://academic.oup.com/mbe/article-abstract/36/6/1281/5420164 by Ghent University user on 11 March 2020
diversity at both population and species levels. Mol Ecol.
26(16):4197–4210.
Smith JM, Haigh J. 1974. The hitch-hiking effect of a favourable gene.
Genet Res. 23(01):23–25.
Supek F, Vlahovicek K. 2004. INCA: synonymous codon usage analysis
and clustering by means of self-organizing map. Bioinformatics
20(14):2329–2330.
Tajima F. 1983. Evolutionary relationship of DNA-sequences in finite
populations. Genetics 105(2):437–460.
Thornton K, Bachtrog D, Andolfatto P. 2006. X chromosomes and auto-
somes evolve at similar rates in Drosophila: no evidence for faster-X
protein evolution. Genome Res. 16(4):498–504.
Torgerson DG, Singh RS. 2006. Enhanced adaptive evolution of sperm-
expressed genes on the mammalian X chromosome. Heredity
96(1):39–44.
Unckless RL, Larracuente AM, Clark AG. 2015. Sex-ratio meiotic drive
and Y-linked resistance in Drosophila affinis.Genetics
199(3):831–840.
Vanthournout B, Busck M, Bechsgaard J, Hendrickx F, Schramm S, Bilde
T. 2018. Male spiders control offspring sex ratio through greater
production of female-determining sperm. Proc R Soc B.
285(1875):20172887.
Vanthournout B, Deswarte K, Hammad H, Bilde T, Lambrecht B,
Hendrickx F. 2014. Flow cytometric sexing of spider sperm reveals
an equal sperm production ratioinafemale-biasedspecies.Biol
Lett. 10:20140159.
Vicoso B, Charlesworth B. 2006. Evolution on the X chromosome: un-
usual patterns and processes. Nat Rev Genet. 7(8):645–653.
Vicoso B, Charlesworth B. 2009. Effective population size and the faster-X
effect: an extended model. Evolution 63(9):2413–2426.
Vindelov LL, Christensen IJ, Nissen NI. 1983. A detergent-trypsin method
for the preparation of nuclei for flow cytometric DNA analysis.
Cytometry 3(5):323–327.
WrightAE,HarrisonPW,ZimmerF,MontgomerySH,PointerMA,Mank
JE. 2015. Variation in promiscuity and sexual selection drives avian
rate of Faster-Z evolution. Mol Ecol. 24(6):1218–1235.
Xu K, Oh S, Park T, Presgraves DC, Yi SV. 2012. Lineage-specific variation
in slow- and fast-X evolution in primates. Evolution 66(6):1751–1761.
Yang ZH. 2007. PAML 4: phylogenetic analysis by maximum likelihood.
Mol Biol Evol. 24(8):1586–1591.
Faster X Chromosome Evolution in Spiders .doi:10.1093/molbev/msz074 MBE
1293
Downloaded from https://academic.oup.com/mbe/article-abstract/36/6/1281/5420164 by Ghent University user on 11 March 2020
... We used additional data to select the loci, such as the location of homologous sequences in the spiders Trichonephila senegalensis (Araneidae) (Walckenaer, 1841) (M. Grewoldt, J. Schneider, and M. Cordellier, unpublished data) and Stegodyphus mimosarum (Eresidae) Pavesi, 1883 (Bechsgaard et al., 2019). We tested the amplification with conventional PCR, checked on a 1.5% agarose gel with ethidium bromide staining and ultraviolet transillumination. ...
Article
Full-text available
Differences between sexes in growth, ecology and behaviour strongly shape species biology. In some animal groups, such as spiders, it is difficult or impossible to identify the sex of juveniles based on external morphology. This information would be useful for field surveys, behavioural experiments and ecological studies, such as those on sex ratios and dispersal. In species with sex chromosomes, sex can be determined based on the specific sex chromosome complement. Additionally, information on the sequence of sex chromosomes provides the basis for studying sex chromosome evolution. We combined cytogenetic and genomic data to identify the sex chromosomes in the sexually dimorphic spider Argiope bruennichi and designed quantitative real-time polymerase chain reaction sex markers. We found that the genome size and GC content of this spider fall into the ranges reported for the majority of araneids. The male karyotype is formed by 24 acrocentric chromosomes with an X1X20 sex chromosome system, with little similarity between X chromosomes, suggesting an origin of these chromosomes by X chromosome fission or early duplication of an X chromosome and subsequent independent differentiation of the copies. Our data suggest X chromosomes of similar sizes in A. bruennichi. They are smaller chromosomes of the complement. Our findings open the door to new directions in spider evolutionary and ecological research.
... mating system, female-biased sex ratio, and frequent extinction-colonization events has severe consequences for population genetic diversity and structure: genetic diversity is extremely low within populations, and despite restricted gene flow between populations, population extinctionrecolonization causes remarkably low genetic divergence among populations (Lubin and Bilde 2007;Vanthournout et al. 2018;Settepani et al. 2014Settepani et al. , 2017Bechsgaard et al. 2019). This type of host metapopulation dynamics may have implications for host-microbiome dynamics (Mihaljevic 2012;Miller et al. 2018), but potentially also for the spatial distribution of the microbiome composition. ...
Article
Full-text available
Host symbiont interactions may form obligatory or facultative associations that are context dependent. Long-term studies on microbiome composition from wild populations should assess the temporal and spatial dynamics of host-microbe associations. We characterized the temporal and spatial variation in the bacterial microbiome composition in six populations of the social spider Stegodyphus dumicola for 2.5 years, using 16S rRNA gene amplicon sequencing of whole spiders. Individuals within a nest exhibit highly similar microbiomes, which remain stable over several generations and are not predictably affected by seasonal variation in temperature or humidity. This stability in nest microbiome is likely due to social transmission, whereas drift-like processes during new nest foundations explain variation in host microbiomes between nests. This is supported by the lack of obligate symbionts (i.e. no symbionts are present in all spider individuals). Quantitative PCR analyses showed that the bacterial load of individual spiders is stable in healthy nests but can increase dramatically in perishing nests. These increases are not driven by specific bacterial taxa but likely caused by loss of host immune control under deteriorating conditions. Spider nests show an annual survival rate of approximately 45%, but nest death is not correlated to microbiome composition, and the bacteria found in S. dumicola are not considered to be high virulence pathogens.
... Therefore, the effective population size and recombination rate for X chromosomes should be lower than autosomes. This leads to the theoretical prediction of increased evolutionary rates of X chromosomes under natural selection, especially in the case of recessive variants (Charlesworth, Coyne & Barton, 1987;Ellegren, 2009;Bechsgaard et al., 2019). In the case of A. bruennichi, it seems that at least one of the X chromosomes has indeed evolved more quickly than autosomes. ...
Preprint
Anthropogenic climate change is rapidly altering ecosystems, driving range shifts, range contractions, dwindling population sizes and local extinctions in many species. Some species, however, are expanding their ranges and seem to benefit from warming temperatures. This is the case for the wasp spider, Argiope bruennichi, which has undergone a range expansion from its historic range in the Mediterranean ("core"), now reaching as far as the Baltic states and Scandinavia ("edge"). The rate of this range expansion cannot be attributed to climate change alone, and it has been hypothesized that adaptive introgression lent the genetic variation upon which selection could act, enabling the rapid range expansion. In the present study, we test this hypothesis, firstly by quantifying the degree of local adaptation and phenotypic plasticity in cold tolerance in edge relative to core populations, then by investigating genomic and phenotypic turnover across the proposed introgression zone. With a reciprocal transplant common garden experiment, we provide strong support for the hypothesis that edge populations are locally adapted to colder winter conditions. We also find evidence of seasonal plasticity in the core populations, while edge populations have lost this plasticity. Our genome-wide analysis, using a combination of FST outlier and genetic-environment association tests, supports the hypothesis that adaptive introgression played a role in environmental adaptation.
... Recent progress in spider genomics has led to publication of first spider genome assemblies [40,41] along with the first sequences linked with X chromosomes [42], but so far, no data about the genetic composition of spider Y chromosomes have been acquired. ...
Article
Full-text available
Spiders are an intriguing model to analyse sex chromosome evolution because of their peculiar multiple X chromosome systems. Y chromosomes were considered rare in this group, arising after neo-sex chromosome formation by X chromosome-autosome rearrangements. However, recent findings suggest that Y chromosomes are more common in spiders than previously thought. Besides neo-sex chromosomes, they are also involved in the ancient X1X2Y system of haplogyne spiders, whose origin is unknown. Furthermore, spiders seem to exhibit obligatorily one or two pairs of cryptic homomorphic XY chromosomes (further cryptic sex chromosome pairs, CSCPs), which could represent the ancestral spider sex chromosomes. Here, we analyse the molecular differentiation of particular types of spider Y chromosomes in a representative set of ten species by comparative genomic hybridisation (CGH). We found a high Y chromosome differentiation in haplogyne species with X1X2Y system except for Loxosceles spp. CSCP chromosomes exhibited generally low differentiation. Possible mechanisms and factors behind the observed patterns are discussed. The presence of autosomal regions marked predominantly or exclusively with the male or female probe was also recorded. We attribute this pattern to intraspecific variability in the copy number and distribution of certain repetitive DNAs in spider genomes, pointing thus to the limits of CGH in this arachnid group. In addition, we confirmed nonrandom association of chromosomes belonging to particular CSCPs at spermatogonial mitosis and spermatocyte meiosis and their association with multiple Xs throughout meiosis. Taken together, our data suggest diverse evolutionary pathways of molecular differentiation in different types of spider Y chromosomes.
... While high gene turnover on X chromosomes has been observed in Dipteran insects (Vicoso and Bachtrog 2015), a study on Hemiptera showed that differentiated sex chromosomes are extremely stable (Pal and Vicoso 2015). Bechsgaard et al. (2019) very recently published the first study on the evolution of X-linked regions in spiders. Their approach made clever use of flow cytometry to sort sperm cells from the social spider Stegodyphus mimosarum into two groups, those containing X chromosomes and those who did not. ...
Article
Full-text available
Sexual reproduction is pervasive in animals and has led to the evolution of sexual dimorphism. In most animals, males and females show marked differences in primary and secondary sexual traits. The formation of sex-specific organs and eventually sex-specific behaviors is defined during the development of an organism. Sex determination processes have been extensively studied in a few well-established model organisms. While some key molecular regulators are conserved across animals, the initiation of sex determination is highly diverse. To reveal the mechanisms underlying the development of sexual dimorphism and to identify the evolutionary forces driving the evolution of different sexes, sex determination mechanisms must thus be studied in detail in many different animal species beyond the typical model systems. In this perspective article, we argue that spiders represent an excellent group of animals in which to study sex determination mechanisms. We show that spiders are sexually dimorphic in various morphological, behavioral, and life history traits. The availability of an increasing number of genomic and transcriptomic resources and functional tools provides a great starting point to scrutinize the extensive sexual dimorphism present in spiders on a mechanistic level. We provide an overview of the current knowledge of sex determination in spiders and propose approaches to reveal the molecular and genetic underpinnings of sexual dimorphism in these exciting animals.
... Relaxed purifying selection is expected to be especially common in species with low effective population size (N e ), and social species such as social spiders are expected to experience low N e as a result of reproductive skew, female-biased sex ratios, and inbreeding (Romiguier et al. 2014;Settepani et al. 2016;Galtier et al. 2018). Indeed, previous studies in social spiders (Settepani et al. 2014(Settepani et al. , 2016(Settepani et al. , 2017Bechsgaard et al. 2019) and also social insects have found evidence for low N e and genomewide relaxed purifying selection when compared with solitary species (Romiguier et al. 2014;Kapheim et al. 2015;Galtier et al. 2018). Future studies using both polymorphism data and divergence data will be necessary to further tease apart the contribution of elevated positive selection and relaxed purifying selection (Yang and Bielawski 2000;Nielsen 2005) to spider genome evolution. ...
Article
Full-text available
Comparative genomics has begun to elucidate the genomic basis of social life in insects but insight into the genomic basis of spider sociality has lagged behind. To begin to characterize genomic signatures associated with the evolution of social life in spiders, we performed one of the first spider comparative genomics studies including five solitary species and two social species, representing two independent origins of sociality in the genus Stegodyphus. We found that the two social spider species had a large expansion of gene families associated with transport and metabolic processes and an elevated genome-wide rate of molecular evolution compared with the five solitary spider species. Genes that were rapidly evolving in the two social species relative to the five solitary species were enriched for transport, behavior, and immune functions, while genes that were rapidly evolving in the solitary species were enriched for energy metabolism processes. Most rapidly evolving genes in the social species S. dumicola were broadly expressed across four tissues and enriched for transport functions, but 12 rapidly evolving genes showed brain-specific expression and were enriched for social behavioral processes. Altogether, our study identifies putative genomic signatures and potential candidate genes associated with spider sociality. These results indicate that future spider comparative genomic studies, including broader sampling and additional independent origins of sociality, can further clarify the genomic causes and consequences of social life.
Article
Although eusocial animals often achieve ecological dominance in the ecosystems where they occur, many populations are unstable, resulting in local extinction. Both patterns may be linked to the characteristic demography of eusocial species—high reproductive skew and reproductive division of labor support stable effective population sizes that make eusocial groups more competitive in some species, but also lower effective population sizes that increase susceptibility to population collapse in others. Here, we examine the relationship between demography and social organization in Synalpheus snapping shrimps, a group in which eusociality has evolved recently and repeatedly. We show using coalescent demographic modelling that eusocial species have had lower but more stable effective population sizes across 100,000 generations. Our results are consistent with the idea that stable population sizes may enable competitive dominance in eusocial shrimps, but they also suggest that recent population declines are likely caused by eusocial shrimps’ heightened sensitivity to environmental changes, perhaps as a result of their low effective population sizes and localized dispersal. Thus, although the unique life histories and demography of eusocial shrimps have likely contributed to their persistence and ecological dominance over evolutionary timescales, these social traits may also make them vulnerable to contemporary environmental change.
Article
Full-text available
Although widely thought to be aggressive, solitary, and potentially cannibalistic, some spider species have evolved group-living behaviors. The distinct transition provides the framework to uncover group-living evolution. Here, we conducted a comparative transcriptomic study and examined patterns of molecular evolution in two independently evolved group-living spiders and twelve solitary species. We report that positively selected genes among group-living spider lineages are significantly enriched in nutrient metabolism and autophagy pathways. We also show that nutrient-related genes of group-living spiders convergently experience amino acid substitutions and accelerated relative evolutionary rates. These results indicate adaptive convergence of nutrient metabolism that may ensure energy supply in group-living spiders. The decelerated evolutionary rate of autophagy-related genes in group-living lineages is consistent with an increased constraint on energy homeostasis as would be required in a group-living environment. Together, the results show that energy metabolic pathways play an important role in the transition to group-living in spiders. 蜘蛛普遍被认为具攻击性、独居、潜在同类相食,但是部分蜘蛛进化出了群居的行为。基于比较转录组研究,发现群居蜘蛛受到正选择基因富集在营养代谢和自噬通路上。结果表明能量代谢通路在改变蜘蛛从攻击性到社会性中的作用。
Article
Full-text available
The faster-X effect, namely the rapid evolution of protein-coding genes on the X-chromosome, has been widely reported in metazoans. However, the prevalence of this phenomenon across diverse systems and its potential causes remain largely unresolved. Analysis of sex-biased genes may elucidate its possible mechanisms: for example, in systems with X/Y males a more pronounced faster-X effect in male-biased genes than in female-biased or unbiased genes may suggest fixation of recessive beneficial mutations rather than genetic drift. Further, theory predicts that the faster-X effect should be promoted by X-chromosome dosage compensation. Here, we asked whether we could detect a faster-X effect in genes of the beetle Tribolium castaneum (and T. freemani orthologs), which has X/Y sex-determination and heterogametic males. Our comparison of protein sequence divergence (dN/dS) on the X-chromosome versus autosomes indicated a rarely observed absence of a faster-X effect in this organism. Further, analyses of sex-biased gene expression revealed that the X-chromosome was particularly highly enriched for ovary-biased genes, which evolved slowly. In addition, an evaluation of male X-chromosome dosage compensation in the gonads and in non-gonadal somatic tissues indicated a striking lack of compensation in the testis. This under-expression in testis may limit fixation of recessive beneficial X-linked mutations in genes transcribed in these male sex organs. Taken together, these beetles provide an example of the absence of a faster-X effect on protein evolution in a metazoan, that may result from two plausible factors, strong constraint on abundant X-linked ovary-biased genes and a lack of gonadal dosage compensation.
Preprint
Full-text available
Background The faster-X effect, namely the rapid evolution of protein-coding genes on the X-chromosome, has been reported in numerous metazoans. However, the prevalence of this phenomenon across metazoans and its potential causes remain largely unresolved. Analysis of sex-biased genes may elucidate its possible mechanisms: a more pronounced faster-X effect in male-biased genes than in female-biased or unbiased genes, suggests fixation of recessive beneficial mutations rather than genetic drift. Further, theory predicts that the faster-X effect should be promoted by X-chromosome dosage compensation, but this topic remains rarely empirically examined. Results Here, we asked whether we could detect a faster-X effect in genes of the beetle Tribolium castaneum (and T. freemani orthologs), which has X/Y sex-determination and heterogametic males. Our comparison of protein sequence divergence (dN/dS) on the X-chromosome versus autosomes indicated the complete absence of a faster-X effect. Further, analyses of sex-biased gene expression revealed that the X-chromosome was strongly enriched for ovary-biased genes, which evolved under exceptionally high constraint. An evaluation of male X-chromosome dosage compensation in the gonads and in non-gonadal somatic tissues showed an extreme lack of compensation in the testis. This under-expression of the X chromosome in males may limit the phenotypic effect, and therefore likelihood of fixation, of recessive beneficial X-linked mutations in genes transcribed in male gonads. Conclusions We show that these beetles display a rare unequivocal example of the absence of a faster-X effect in a metazoan. We propose two potential causes for this, namely high constraint on X-linked ovary-biased genes, and an extreme lack of dosage compensation of genes transcribed in the testis.
Article
Full-text available
The faster evolution of X chromosomes has been documented in several species, and results from the increased efficiency of selection on recessive alleles in hemizygous males and/or from increased drift due to the smaller effective population size of X chromosomes. Aphids are excellent models for evaluating the importance of selection in faster-X evolution because their peculiar life cycle and unusual inheritance of sex chromosomes should generally lead to equivalent effective population sizes for X and autosomes. Because we lack a high-density genetic map for the pea aphid, whose complete genome has been sequenced, we first assigned its entire genome to the X or autosomes based on ratios of sequencing depth in males (X0) to females (XX). Then, we computed non-synonymous to synonymous substitutions ratios (dN/dS) for the pea aphid gene set and found faster evolution of X-linked genes. Our analyses of substitution rates, together with polymorphism and expression data, showed that relaxed selection is likely to be the greatest contributor to faster-X because a large fraction of X-linked genes are expressed at low rates and thus escape selection. Yet, a minor role for positive selection is also suggested by the difference between substitution rates for X and autosomes for male-biased genes (but not for asexual female-biased genes) and by lower Tajima's D for X-linked compared to autosomal genes with highly male-biased expression patterns. This study highlights the relevance of organisms displaying alternative chromosomal inheritance to the understanding of forces shaping genome evolution.
Article
Full-text available
Division of reproductive behaviour and alloparental care are key aspects of many animal societies. In cooperatively breeding species, variation in helping effort and unequal task participation are frequently observed. However, the extent to which the reproductive state of an individual affects the tasks performed during offspring care remains poorly understood. In the social spider Stegodyphus dumicola, approximately 40% of females reproduce, and mothers show extended maternal care including eggsac tending, regurgitation feeding and matriphagy, in which they are consumed by the offspring. We asked whether and to what extent virgin females participate in extreme maternal care and whether they differ from reproducing females in foraging activity. We show that virgin females contributed to all aspects of extended brood care, including regurgitation feeding and matriphagy. This suggests a physiological adaptation in virgin females to cooperative breeding, since in the subsocial Stegodyphus lineatus only mated females provide extended maternal care. Although virgin females and mothers are behaviourally totipotent, we found evidence for task differentiation as virgins engaged less in brood care and more in prey attack than mothers. High relatedness among nestmates and low probability of future reproduction in virgin helpers suggest alignment of reproductive interests between mothers and allomothers. Therefore, extreme allomaternal care by virgin helpers can be considered an adaptation to cooperative breeding in social spiders.
Article
Sex allocation theory predicts that when sons and daughters have different reproductive values, parents should adjust offspring sex ratio towards the sex with the higher fitness return. Haplo-diploid species directly control offspring sex ratio, but species with chromosomal sex determination (CSD) were presumed to be constrained by Mendelian segregation. There is now increasing evidence that CSD species can adjust sex ratio strategically, but the underlying mechanism is not well understood. One hypothesis states that adaptive control is more likely to evolve in the heterogametic sex through a bias in gamete production. We investigated this hypothesis in males as the heterogametic sex in two social spider species that consistently show adaptive female-biased sex ratio and in one subsocial species that is characterized by equal sex ratio. We quantified the production of male (0) and female (X) determining sperm cells using flow cytometry, and show that males of social species produce significantly more X-carrying sperm than 0-sperm, on average 70%. This is consistent with the production of more daughters. Males of the subsocial species produced a significantly lower bias of 54% X-carrying sperm. We also investigated whether inter-genomic conflict between hosts and their endosymbionts may explain female bias. Next generation sequencing showed that five common genera of bacterial endosymbionts known to affect sex ratio are largely absent, ruling out that endosymbiont bacteria bias sex ratio in social spiders. Our study provides evidence for paternal control over sex allocation through biased gamete production as a mechanism by which the heterogametic sex in CSD species adaptively adjust offspring sex ratio.
Article
A faster rate of adaptive evolution of X-linked genes compared with autosomal genes can be caused by the fixation of recessive or partially recessive advantageous mutations, due to the full expression of X-linked mutations in hemizygous males. Other processes, including recombination rate and mutation rate differences between X chromosomes and autosomes, may also cause faster evolution of X-linked genes. We review population genetics theory concerning the expected relative values of variability and rates of evolution of X-linked and autosomal DNA sequences. The theoretical predictions are compared with data from population genomic studies of several species of Drosophila. We conclude that there is evidence for adaptive faster-X evolution of several classes of functionally significant nucleotides. We also find evidence for potential differences in mutation rates between X-linked and autosomal genes, due to differences in mutational bias towards GC to AT mutations. Many aspects of the data are consistent with the male hemizygosity model, although not all possible confounding factors can be excluded.
Article
The effective population size (Ne) is a central factor in determining maintenance of genetic variation. The neutral theory predicts that loss of variation depends on Ne, with less genetic drift in larger populations. We monitored genetic drift in 42 Drosophila melanogaster populations of different adult census population sizes (10, 50 or 500) using pooled RAD sequencing. In small populations, variation was lost at a substantially lower rate than expected. This observation was consistent across two ecological relevant thermal regimes, one stable and one with a stressful increase in temperature across generations. Estimated ratios between Ne and adult census size were consistently higher in small than in larger populations. The finding provides evidence for a slower than expected loss of genetic diversity and consequently a higher than expected long-term evolutionary potential in small fragmented populations. More genetic diversity was retained in areas of low recombination, suggesting that associative overdominance, driven by disfavored homozygosity of recessive deleterious alleles, is responsible for the maintenance of genetic diversity in smaller populations. Consistent with this hypothesis, the X-chromosome, which is largely free of recessive deleterious alleles due to hemizygosity in males, fits neutral expectations even in small populations. Our experiments provide experimental answers to a range of unexpected patterns in natural populations, ranging from variable diversity on X-chromosomes and autosomes to surprisingly high levels of nucleotide diversity in small populations. This article is protected by copyright. All rights reserved.
Article
Across several animal taxa, the evolution of sociality involves a suite of characteristics, a 'social syndrome', that includes cooperative breeding, reproductive skew, primary female biased sex-ratio, and the transition from outcrossing to inbreeding mating system, factors that are expected to reduce effective population size (Ne). This social syndrome may be favoured by short-term benefits but come with long-term costs, because the reduction in Ne amplifies loss of genetic diversity by genetic drift, ultimately restricting the potential of populations to respond to environmental change. To investigate the consequences of this social life form on genetic diversity, we used a comparative RAD-sequencing approach to estimate genome-wide diversity in spider species that differ in level of sociality, reproductive skew, and mating system. We analysed multiple populations of three independent sister-species pairs of social inbreeding and subsocial outcrossing Stegodyphus spiders, and a subsocial outgroup. Heterozygosity and within population diversity were 6-10 fold lower in social compared to subsocial species, and demographic modelling revealed a tenfold reduction in Ne of social populations. Species-wide genetic diversity depends on population divergence and the viability of genetic lineages. Population genomic patterns were consistent with high lineage turnover, which homogenizes the genetic structure that builds up between inbreeding populations, ultimately depleting genetic diversity at the species level. Indeed, species-wide genetic diversity of social species was 5-8 times lower than that of subsocial species. The repeated evolution of species with this social syndrome is associated with severe loss of genome-wide diversity, likely to limit their evolutionary potential. This article is protected by copyright. All rights reserved.