David et al. BMC Genomics 2010, 11:216
http://www.biomedcentral.com/1471-2164/11/216Open AccessR E S E A R C H A R T I C L E
Research articleTranscriptome response to pollutants and
insecticides in the dengue vector Aedes aegypti
using next-generation sequencing technology
Jean-Philippe David*, Eric Coissac, Christelle Melodelima, Rodolphe Poupardin, Muhammad Asam Riaz,
Alexia Chandor-Proust and Stéphane Reynaud
Abstract
Background: The control of mosquitoes transmitting infectious diseases relies mainly on the use of chemical
insecticides. However, mosquito control programs are now threatened by the emergence of insecticide resistance.
Hitherto, most research efforts have been focused on elucidating the molecular basis of inherited resistance. Less
attention has been paid to the short-term response of mosquitoes to insecticides and pollutants which could have a
significant impact on insecticide efficacy. Here, a combination of LongSAGE and Solexa sequencing was used to
perform a deep transcriptome analysis of larvae of the dengue vector Aedes aegypti exposed for 48 h to sub-lethal
doses of three chemical insecticides and three anthropogenic pollutants.
Results: Thirty millions 20 bp cDNA tags were sequenced, mapped to the mosquito genome and clustered,
representing 6850 known genes and 4868 additional clusters not located within predicted genes. Mosquitoes exposed
to insecticides or anthropogenic pollutants showed considerable modifications of their transcriptome. Genes
encoding cuticular proteins, transporters, and enzymes involved in the mitochondrial respiratory chain and
detoxification processes were particularly affected. Genes and molecular mechanisms potentially involved in
xenobiotic response and insecticide tolerance were identified.
Conclusions: The method used in the present study appears as a powerful approach for investigating fine
transcriptome variations in genome-sequenced organisms and can provide useful informations for the detection of
novel transcripts. At the biological level, despite low concentrations and no apparent phenotypic effects, the
significant impact of these xenobiotics on mosquito transcriptomes raise important questions about the 'hidden
impact' of anthropogenic pollutants on ecosystems and consequences on vector control.
Background
During the past 60 years, the amount of anthropogenic
xenobiotics released into natural ecosystems has dramati-
cally increased. Although the effect of these chemicals on
human health is intensively studied, their impact on other
organisms remains poorly understood. Because pollut-
ants often accumulate in fresh-water bodies and sedi-
ments [1], their impact on wetland fauna is of importance
for these ecosystems. Among aquatic arthropods found
in wetlands, mosquitoes are distributed worldwide and
are often exposed to anthropogenic pollutants and insec-
ticides during their aquatic larval stage. Indeed insecti-
cides are often deliberately introduced into the mosquito
habitat in the fight against the many human diseases they
transmit (e.g. malaria, dengue fever, yellow fever and
filariasis) [2]. As a consequence mosquito control pro-
grams are now threatened by the selection of mosquito
populations resistant to these chemical insecticides [3].
Differential gene transcription in insecticide-resistant
mosquitoes has been frequently used to identify genes
putatively involved in inherited metabolic resistance
mechanisms [4-7]. For that purpose most approaches
used cDNA microarrays and were often focused on genes
encoding enzymes potentially involved in the bio-trans-* Correspondence: jean-philippe.david@ujf-grenoble.fr
1BioMed Central
© 2010 David et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons
Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.
formation of insecticides molecules [8,9], although recent
findings suggest that the differential expression of other
Laboratoire d'Ecologie Alpine (LECA, UMR 5553 CNRS - Université Grenoble),
France
Full list of author information is available at the end of the article

David et al. BMC Genomics 2010, 11:216
http://www.biomedcentral.com/1471-2164/11/216
Page 2 of 12
transcripts may also contribute to insecticide tolerance
[4,10]. Less attention has been paid to the short term
transcriptome response of insects to xenobiotics, though
this may lead to the discovery of novel molecular mecha-
nisms contributing to insecticide tolerance [11-13]. We
recently demonstrated that exposing mosquito larvae to
low concentrations of pollutants for a few hours can
increase their tolerance to chemical insecticides, possibly
due to an alteration of the expression of detoxification
enzymes [11,12]. In this context, understanding cross
responses of mosquitoes to insecticides and pollutants at
the whole transcriptome level may ultimately lead to
improvements in vector control strategies by optimizing
insecticide treatments in polluted areas [7]. Moreover,
deciphering transcriptome response of mosquitoes to
anthropogenic xenobiotics may identify genes involved in
chemical stress response that were not detected by stan-
dard toxicological studies.
Today, quantitative transcriptomic methods are diversi-
fied and divided into two kind of technology: 'closed' and
'open' techniques depending on genome annotation con-
straints [14,15]. In 'closed' technologies, gene expression
microarrays are the standard method used for transcrip-
tome analysis. However, this type of technology does not
allow the characterization and analysis of new transcripts
and suffers from various technical biases such as non-
specific hybridization and insufficient signal for low
expressed genes. In contrast, 'open' transcriptome analy-
ses based on the sequencing of either ESTs or short
cDNA tags, like Serial Analysis of Gene Expression
(SAGE) [16], LongSAGE [17] and Massive Parallel Signa-
ture Sequencing (MPSS) [18] can measure the transcript
level of both known and unknown genes [19]. The short
cDNA tags obtained by LongSAGE or MPSS can directly
be mapped to the genome sequence, allowing the identifi-
cation of new transcripts [15]. Because these sequencing
techniques do not target a defined portion of cDNAs,
these approaches are not optimized for the deep analysis
of transcriptome variations [20]. Recently, a combination
of LongSAGE and Solexa sequencing technology, leading
to the production and sequencing of millions of tags on a
defined region of cDNAs, has been used to characterize
mouse hypothalamus transcriptome [15]. To our knowl-
edge, this new method, called Digital Gene Expression
Tag Profiling (DGETP) has never been used to compare
whole transcriptome variations of a non-mammalian
organism in different environmental conditions.
Here, we used the DGETP approach to perform a deep
transcriptome analysis of larvae of the mosquito Aedes
aegypti exposed to different anthropogenic xenobiotics.
We examined the effect of sublethal doses of three pollut-
used for mosquito control (the pyrethroid permethrin,
the neonicotinoid imidacloprid and the carbamate
propoxur). This approach was suitable for investigating
deep transcriptome variations in mosquitoes and identi-
fied several loci with high transcription signal not previ-
ously identified in mosquito genome. At the biological
level, the transcript levels of many genes were affected by
xenobiotic exposure. Several genes and protein families
responding to individual or multiple xenobiotics were
identified, unraveling the complexity of xenobiotic-
response in mosquitoes and identifying genes potentially
involved in insecticide tolerance or biological interactions
between insecticides and pollutants.
Results
Sequencing, mapping and clustering of cDNA tags
By sequencing 7 cDNA tag libraries from mosquito larvae
exposed to different xenobiotics, a total of 29.45 million
reads (100% of total reads) corresponding to 726,269 dis-
tinct 20-mer tags were obtained (Table 1). By removing
any tag represented by less than 20 reads across all librar-
ies, background filtering slightly reduced the total num-
ber of reads to 28.12 million (95.5%) but greatly reduced
the number of distinct tags to 33,037. Among them,
15,253 distinct tags were successfully mapped onto the
Ae. aegypti genome at a unique genomic location without
mismatch, representing 15.2 million reads (51.6%).
Among successfully mapped tags, 9,812 distinct tags
(12.59 million reads, 42.7%) were mapped to 6,850 pre-
dicted genes while the remaining reads (8.9%) were
mapped outside gene boundaries (see methods).
Clustering analysis of 20-mer cDNA tags successfully
mapped to mosquito genome allowed us to identify a
total of 13,118 distinct clusters including 8,250 clusters
associated to predicted genes. Distribution of the total
number of reads across genes, clusters and tags (Addi-
tional file 1: Suppl. Figure 1) spanned more than 4 orders
of magnitude with most genes/clusters being represented
by 25 to 5000 reads. Median total number of reads per
gene, cluster, tag and cluster not mapped within pre-
dicted gene were 217, 124, 101 and 79 respectively.
Quantitative transcription data obtained from cDNA tags
Analysis of transcription levels in mosquito larvae
exposed to each xenobiotic was performed at the gene
level for tags mapped within predicted genes (i.e. gather-
ing all tags mapped within each gene) and at the cluster
level for tags not mapped within predicted genes (i.e.
gathering all tags mapped within each cluster). This anal-
ysis identified 453 genes and 225 additional clusters with
a mean transcript ratio (TR) significantly > 2-fold inants likely to be found in wetlands (the herbicide atrazine,
the polycyclic aromatic hydrocarbon fluoranthene and
the heavy metal copper) and three chemical insecticides
either direction in at least 1 condition (Fisher's test Pvalue
< 10-3 after multiple testing correction). Overall distribu-
tion of TRs and their associated Pvalues revealed a well-bal-

David et al. BMC Genomics 2010, 11:216
http://www.biomedcentral.com/1471-2164/11/216
Page 3 of 12
anced distribution between over- and under transcription
with TRs ranging from 600-fold under transcription to
more than 2000-fold over transcription compared with
controls (Figure 1 and Additional file 2: Suppl. Table 1).
Cross-validation of TRs with real-time quantitative RT-
PCR on 14 genes (Additional file 3: Suppl. Figure 2)
revealed a good correlation of TRs obtained from the two
techniques (r = 0.71 and P = 4.16 E-05), although the
DGETP method often produced higher TRs (in either
direction) than real-time quantitative RT-PCR.
Overall transcriptome variations across treatments
Global analysis of transcriptome variations between mos-
quito larvae exposed to each xenobiotic revealed that the
proportion of genes/clusters differently transcribed var-
ied greatly between treatments (Table 2). This proportion
ranged from 0.26% to 3.94% of all detected genes/clusters
for permethrin and propoxur respectively. No correlation
was found between the number of genes/clusters differ-
entially transcribed in each treatment and the number of
reads sequenced or the number of cDNA tags success-
fully mapped to genome, suggesting an accurate normal-
ization across all libraries. When considering organic
xenobiotics (all but copper), the number of genes/clusters
differentially transcribed for each treatment was signifi-
cantly positively correlated with the molarity of the xeno-
biotic used for larval exposure, (r = 0.89 and P < 0.05).
This overall positive correlation revealed that despite the
different nature of xenobiotics, increasing the number of
organic molecules lead to an increase in the number of
genes/cluster differentially transcribed. Principal compo-
nent analysis (PCA) based on TRs of genes/clusters dif-
ferentially transcribed revealed similar transcriptome
variations of mosquito larvae exposed to the two chemi-
cal insecticides propoxur and imidacloprid and the poly-
cyclic aromatic hydrocarbon fluoranthene (Additional
file 4: Suppl. Figure 3). Conversely, transcriptome varia-
tions of larvae exposed to the insecticide permethrin, the
herbicide atrazine and copper were more specific.
Genes differentially transcribed across treatments
Functional analysis of the 453 genes differentially tran-
scribed in mosquito larvae exposed to xenobiotics
revealed that genes responding to xenobiotics encode
Table 1: Sequencing statistics
Reads Ctrl
(×106)
Copper
(×106)
Fluo
(×106)
Atraz
(×106)
Propo
(×106)
Perm
(×106)
Imida
(×106)
Mean
(×106)
Total
(×106)
% Total Distinct
tags
Sequenced 4.35 4.30 4.41 2.75 3.88 4.90 4.85 4.21 29.45 100 726 269
Filtered
from
background
4.16 4.10 4.21 2.63 3.72 4.68 4.62 4.02 28.12 95.5 33 037
Mapped to
genome
2.27 2.31 2.29 1.42 1.80 2.63 2.48 2.17 15.20 51.6 15 253
Mapped to
genes
1.89 1.93 1.87 1.19 1.49 2.19 2.03 1.80 12.59 42.7 9 812
Reads filtered from background represent tags showing > 20 reads across all conditions. Reads mapped to genome represent tags mapped to a
unique genomic location without mismatch. Reads mapped to genes represent tags filtered from background and mapped to predicted genes.
Ctrl: controls; Copper: exposed to copper sulfate; Fluo: exposed to fluoranthene; Atraz: exposed to atrazine; Propo: exposed to propoxur; Perm:
exposed to permethrin; Imida: exposed to imidacloprid.
Figure 1 Distribution and significance of transcription variations
in mosquito larvae exposed to xenobiotics. Transcription ratios of
genes are shown as black dots while genomic clusters not mapped
within genes are shown as white dots. Differential transcription is indi-
cated as a function of both log10 transcription ratios (exposed to xeno- proteins with diverse functions, including a large propor-
tion (up to 50%) of proteins of unknown function (Figure
2 and Additional file 1: Suppl Table 1). Among them, 108
biotics/controls) and Fisher's test Pvalues. Only the transcription ratios of
453 genes and 250 clusters showing a Fisher's test Pvalue < 0.001 in at
least one condition are shown.

David et al. BMC Genomics 2010, 11:216
http://www.biomedcentral.com/1471-2164/11/216
Page 4 of 12
genes were affected by both pollutants and insecticides.
Several genes affected by xenobiotics encoded enzymes,
cuticular proteins and proteins involved in transport or
DNA interactions. As previously shown by PCA, the two
chemical insecticides propoxur and imidacloprid, and to
a lesser extent the polycyclic hydrocarbon fluoranthene,
induce similar functional responses. Response induced by
copper appeared distinct compared to organic xenobiot-
ics, with a high proportion of enzymes being over tran-
scribed. Conversely, response to organic xenobiotics was
characterized by the overproduction of a large proportion
of transcripts encoding cuticular proteins. For these com-
pounds, a positive correlation was found between their
lipophilicity (Log Kow) and the proportion of transcripts
encoding cuticular proteins being significantly over-pro-
duced (r = 0.91; P < 0.01; Log Kow from 0.57 for imidaclo-
prid to 6.1 for permethrin,). Genes encoding cytoskeleton
and ribosomal proteins were also affected by various xen-
obiotics with cytoskeleton proteins showing a marked
repression in larvae exposed to the herbicide atrazine.
Finally, genes encoding proteins involved in transport
were also differentially affected by xenobiotics. A nega-
tive correlation was found between the lipophilicity (Log
Kow) of organic xenobiotics and the number of tran-
scripts involved in transport being over-produced (r =
0.95, P < 0.01).
Impact of xenobiotics on transcripts encoding enzymes
Clustering analysis of genes encoding enzymes signifi-
cantly differentially transcribed in larvae exposed to xen-
obiotics revealed that the transcript level of 115 enzymes
was affected by one or more xenobiotic (Figure 3). The
transcript level of these enzymes was strongly affected in
larvae exposed to the insecticides propoxur and imida-
cloprid and the aromatic hydrocarbon fluoranthene. A
gene tree based on transcript levels across all treatments
revealed a distribution in 6 main different enzyme clus-
ters mainly influenced by these 3 xenobiotics. Twelve
genes encoding enzymes potentially involved in xenobi-
otic detoxification were found differentially transcribed,
including 5 cytochrome P450s monooxygenases (P450s),
4 glutathione S-transferases (GSTs) and 3 carboxy/cho-
linesterases (CCEs). Among them, the three P450s
CYP9M9 (AAEL001807), CYP325X2 (AAEL005696) and
CYP6M11 (AAEL009127) were induced by multiple xen-
obiotics. Interestingly, the cytochrome b5 (AAEL012636),
a co-factor associated with P450 detoxification systems,
was also strongly induced in mosquito larvae exposed to
insecticides and copper. Among GSTs, GSTX2
(AAEL010500) was strongly and specifically induced by
the insecticide propoxur while the induction of GSTD4
encoding enzymes involved in the production of energy
within the respiratory chain such as NADH dehydroge-
nase and ATP synthase were over-produced in mosquito
larvae exposed to xenobiotics while multiple serine pro-
teases, amylases and peptidases were down-regulated.
Discussion
Analyzing transcriptome variations using digital gene
expression tag profiling
Following the genome sequencing of the dengue vector
Ae. aegypti, 15,419 putative genes were identified and
transcripts were detected for 12,350 genes by combining
cDNA microarray, massive parallel signature sequencing
(MPSS) or EST sequencing on several mosquito life
stages [21]. By using the DGETP method, we sequenced
29.4 millions 20-mer tags across 7 distinct cDNA libraries
obtained from 4th-stage larvae. This approach allowed us
to detect significant transcription signals for 6,850 pre-
dicted genes. Considering that several genes may not be
transcribed in 4th-stage larvae and that transcripts
assayed by the DGETP method require the presence of a
DpnII restriction site, such transcriptome coverage
appears satisfactory. Besides, sequence variations
between the Ae. aegypti strain used in our study (Bora-
Bora strain) and the one used for genome sequencing
(Liverpool strain), led to the rejection of numerous reads.
Within our mosquito strain, allelic variations were
detected for numerous loci and also led to the rejection of
a considerable proportion of reads as only alleles exactly
matching to the reference genome sequence were consid-
ered in the analysis (see methods). However, we believe
that such high mapping stringency is critical for generat-
ing accurate gene transcription data with short cDNA
tags. Improving the number of reads by replicating
sequencing libraries for each sample will allow a better
assessment of biological and technical variations together
with increasing transcriptome coverage. By sequencing
10 million random 36 bp cDNA fragments from two
cDNA libraries of females Drosophila melanogaster,
Sackton et al. detected 2,540 annotated genes [22]. By tar-
geting a defined region of cDNAs, the DGETP method
can generate wider transcriptome coverage together with
a higher number of cDNA tags per gene, leading to more
precise gene transcription data. Provided a reference
genome is available and the aim is to quantify transcript
levels between different biological samples, we confirm
that methods based on the combination of LongSAGE
and next-generation sequencing technologies are per-
fectly suited for deep transcriptome analysis [15]. Recent
improvements in sequencing technologies (~30 million
reads/lane on the illumina Genome Analyzer system) are(AAEL001054) appeared less specific. Transcripts encod-
ing esterases were mostly found under produced follow-
ing xenobiotic exposure. Finally, several transcripts
now making sequencing-based approaches the methods
of choice for whole transcriptome analyses.

David et al. BMC Genomics 2010, 11:216
http://www.biomedcentral.com/1471-2164/11/216
Page 5 of 12
Among the 15,253 20-mer cDNA tags successfully
mapped to Ae. aegypti genome, 35% were not located
within predicted gene boundaries extended by 300 bp at
their 3' end (see methods). These tags could be gathered
into 4,868 genomic clusters with more than 40% of them
showing significant transcription signal (> 100 reads,
Additional file 1: Suppl. Figure 1). These clusters may
represent genes, exons or UTR extensions not predicted
by automated annotation. Recent studies revealed that
the genome of complex organisms produce large num-
bers of regulatory noncoding RNAs (ncRNAs) that can be
antisense, intergenic, interleaved or overlapping with
protein-coding genes [23,24]. In that concern, it is likely
that a significant proportion of transcript signatures
detected outside predicted genes represent ncRNAs. The
use of next-generation sequencing approaches specifi-
cally targeting insect ncRNAs will help decipher their
role in mosquito gene regulation and in the capacity of
insects to adapt to different environmental conditions.
Impact of xenobiotics on mosquito larvae transcriptome
to modifications of their chemical environment. The
number of transcripts affected varies greatly depending
on the xenobiotic used for exposure. When considering
organic xenobiotics (all but copper), this number
increased together with the molarity of the xenobiotics.
Our results also revealed that the lipophilicity of the xen-
obiotics affects the number of differentially transcribed
genes encoding cuticular proteins and transporters. It has
been demonstrated that lipophilic xenobiotics accumu-
late in biological membranes or lipid reserves, modifying
their distribution across tissues and cells [25,26].
Although our experimental design did not allow segregat-
ing between the quantity of xenobiotic and their inherent
chemical properties, it is likely that molarity and lipophi-
licity are key factors affecting the magnitude and the
specificity of transcriptome variations observed here.
Our results demonstrated the similar strong transcrip-
tome response of mosquito larvae exposed to the insecti-
cides propoxur and imidacloprid. Despite belonging to
two different chemical groups, the carbamate propoxur
and the neonicotinoid imidacloprid both potentiate the
Table 2: Genes and clusters differentially transcribed after xenobiotic exposure
Genes/
clusters
differentially
transcribed
Copper Fluo Atraz Propo Perm Imida
N % N % N % N % N % N %
Total genes
and additional
clusters
71 0.61 141 1.20 98 0.84 462 3.94 31 0.26 361 3.08
Total genes 49 0.72 86 1.26 60 0.88 318 4.64 20 0.29 239 3.49
Over-
transcribed
46 0.67 50 0.73 25 0.36 130 1.90 16 0.23 113 1.65
Under-
transcribed
3 0.04 36 0.53 35 0.51 188 2.74 4 0.06 126 1.84
Total
additional
clusters not
within genes
22 0.45 55 1.13 38 0.78 144 2.96 11 0.23 122 2.51
Over-
transcribed
18 0.37 36 0.74 21 0.43 53 1.09 9 0.18 51 1.05
Under-
transcribed
4 0.08 19 0.39 17 0.35 91 1.87 2 0.04 71 1.46
For each treatment, the number (N) of genes and additional clusters not mapped within predicted genes found significantly differentially
transcribed are indicated. For each value, the associated percentage regarding the total number of genes (6850), the total number of clusters
not mapped within predicted genes (4868), or the total of genes and additional clusters (11718) is indicated. Genes or clusters were
considered significantly differentially transcribed comparatively to controls if their associated P value (Fisher's test) was < 0.001 after multiple
testing corrections. Copper: exposed to copper sulfate; Fluo: exposed to fluoranthene; Atraz: exposed to atrazine; Propo: exposed to
propoxur; Perm: exposed to permethrin; Imida: exposed to imidacloprid.Global analysis of transcriptome variations associated
with a 48 h exposure of mosquito larvae to low doses of
insecticides and pollutants revealed their ability to adjust
functioning of nicotinic cholinergic receptors [27].
Although genes encoding the primary targets of these
insecticides (acetylcholinesterase or nicotinic receptors)

David et al. BMC Genomics 2010, 11:216
http://www.biomedcentral.com/1471-2164/11/216
Page 6 of 12
were not found significantly differentially transcribed, the
similar transcriptome responses to these two insecticides
may be partly related to similar effects generated by the
alteration of cholinergic neurons functioning [28,29].
We previously demonstrated that exposing mosquito
larvae to various pollutants for few hours can increase
their tolerance to insecticides possibly through an induc-
tion of detoxification enzymes [11,12,30]. Among the dif-
ferent pollutants tested, polycyclic aromatic
hydrocarbons were often the most potent for increasing
insecticide tolerance, possibly due to their ability to
induce detoxification enzymes [31]. The present study
detected a considerable number of genes encoding detox-
ification enzymes (89 cytochrome P450s, 22 GSTs and 27
carboxylesterases) including several genes showing tran-
scription level variations. However, only a small propor-
tion of them were found significantly affected by
xenobiotic exposure, probably due to insufficient number
of reads regarding our Fisher's t test Pvalue threshold.
Among them, members of cytochrome P450 families fre-
quently involved in resistance to insecticides and plant
toxins [7-9,32-34] were over transcribed following expo-
sure to fluoranthene, propoxur or imidacloprid. By
revealing that several other genes with a broad range of
biological functions are similarly affected by insecticides
and pollutants, our results suggest that the impact of pol-
lutants on the ability of mosquitoes to better tolerate
chemical insecticides might also be the consequence of
the induction/repression of other proteins involved in a
wide range of functions. In this concern, several cuticular
from insecticides by cuticular protein thickening leading
to a reduction of insecticide penetration [4,35]. Other
studies demonstrated that cuticular component deposi-
tion is stimulated by environmental stress [36].
Our results also suggest that mosquito larvae exposed
to xenobiotics undertake a metabolic stress associated
with changes of their chemical environment. Global cel-
lular stress response has been defined as all proteins over-
produced due to environmental stress. This response ini-
tially named 'general adaptation syndrome' occurs
together with increased mobilization of energy from stor-
age tissues [37]. Such stress response has been described
for numerous stress factors including exposure to pollut-
ants [38]. In insect cells, response to environmental
aggressions can involve various proteins including heat
shock proteins [39], metallothioneins [40] or p-glycopro-
tein synthesis [41]. Although differentiating between xen-
obiotic-specific and general stress responses is difficult,
we also highlighted such protein families including chap-
eronins, heat shock proteins and ATP-binding cassette
transporters (p-glycoprotein family). Moreover, numer-
ous genes encoding enzymes involved in the production
of energy or in cellular catabolism such as NADH dehy-
drogenase, ATP synthase, trypsin and lipases were found
over transcribed in mosquito larvae exposed to xenobiot-
ics, confirming a global stress response [37,42].
Significant transcript level variations were observed in
response to anthropogenic pollutants though those com-
pounds were not toxic for mosquito larvae (see methods).
Although we predicted the relatively important effect of
Figure 2 Genes differentially transcribed in mosquito larvae exposed to xenobiotics. Analysis was performed on 453 genes found significantly
differentially transcribed in at least 1 condition (Fisher's test Pvalue < 0.001). Genes were assigned to 9 different categories according to their putative
function: enzymes (dark blue), kinases (blue), transport (pink), DNA interaction (purple), cuticle (orange), cytoskeleton dark green), ribosomes (green),
others (grey) and unknown hypothetical proteins (dark grey). For each condition, numbers of genes found significantly over transcribed (A) and under-
transcribed (B) were compared. Copper: exposed to copper sulfate; Fluo: exposed to fluoranthene; Atraz: exposed to atrazine; Propo: exposed to
propoxur; Perm: exposed to permethrin; Imida: exposed to imidacloprid.proteins were found over transcribed in mosquito larvae
exposed to insecticides or organic xenobiotics. It has
been suggested that mosquito may protect themselves
the polycyclic aromatic hydrocarbon (PAH) fluoranthene
on mosquito larvae due to known cellular effects on ani-
mals [11,12,31,43], responses to atrazine and copper were

David et al. BMC Genomics 2010, 11:216
http://www.biomedcentral.com/1471-2164/11/216
Page 7 of 12
Figure 3 Enzymes differentially transcribed in mosquito larvae exposed to xenobiotics. Hierarchical clustering analysis based transcription lev-
els was performed on 115 enzyme-encoding genes showing significant differential transcription (Fisher's test Pvalue < 0.001) in larvae exposed to any
xenobiotic. Gene tree (left) and condition tree (top) were obtained using Pearson's uncentered distance metric calculated from all Log10 transcription
ratios (xenobiotic exposed/controls). Color scale from blue to yellow indicates Log10 transcription ratios from -1 (10-fold under transcription) to +1 (10-
fold over transcription). For each gene, accession number and annotation are indicated. Copper: exposed to copper sulfate; Fluo: exposed to fluoran-
thene; Atraz: exposed to atrazine; Propo: exposed to propoxur; Perm: exposed to permethrin; Imida: exposed to imidacloprid.

David et al. BMC Genomics 2010, 11:216
http://www.biomedcentral.com/1471-2164/11/216
Page 8 of 12
unanticipated. In animals, the cellular impact of PAHs
has been associated with the uncoupling of mitochon-
drial respiration, direct genotoxic damages and the for-
mation of reactive oxygen species [31,44-46]. The over
transcription of NADH dehydrogenase and ATP synthase
observed after exposing larvae to fluoranthene confirm
that similar effects occur in mosquitoes. Although mos-
quitoes do not possess the protein targeted by the triazine
herbicide atrazine (plastoquinone-binding protein in
photosystem II) [47] and a very low concentration was
used (10 μg/L), this chemical affected the transcription of
several mosquito genes. In plants, atrazine disrupts the
electron transport in chloroplasts [48]. In mosquito lar-
vae, several members of the oxidative phosphorylation
pathway including NADH dehydrogenase and ATP syn-
thase were induced by atrazine, suggesting a compensa-
tion for partial uncoupling of oxidative phosphorylation
[44]. Larvae exposed to copper sulfate exhibited a signifi-
cant over transcription of 45 genes including a large pro-
portion of enzymes while only 3 genes were under-
transcribed. The induction of enzymes by copper might
be the consequence of chemical interactions between
Cu2+ ions and metalloenzymes together with other metal-
loproteins involved in electron transfers, hydrolysis and
oxido-reductions [49-51]. The strong induction of the
hemo-protein cytochrome b5 (co-factor of P450s for
electron transfer) together with several serine proteases
and oxidase/peroxidases support this hypothesis.
Conclusions
Overall, despite low concentrations, short exposure time
and no apparent phenotypic modification, the significant
effect of pollutants and insecticides on mosquito larvae
transcriptome raise important questions about the 'hid-
den impact' of anthropogenic pollutants on ecosystems,
including mammals. This concern may even be underes-
timated considering the complex and unknown cross-
effects generated by pollutant mixtures often encoun-
tered in polluted ecosystems [52]. In nematodes, it has
been shown that by applying a realistic heat stress to both
uncontaminated and polluted systems, the specimen
from polluted environment showed a stronger response
[53]. Such effects are likely to occur in polluted mosquito
breeding sites and are likely to affect the efficacy of chem-
ical insecticides used for mosquito control
[4,5,7,11,12,53]. Although further experiments are
required to fully characterize the molecular mechanisms
by which pollutants affect insecticide tolerance in mos-
quitoes, the present study clearly demonstrate that simi-
lar response mechanisms are activated by pollutants and
insecticides. Finally, the persistent contamination of wet-
selection of insecticide resistance mechanisms. Addi-
tional experiments combining exposure of mosquitoes to
pollutants and their subsequent selection with insecti-
cides will provide valuable biological material to answer
this question and may later allow improving mosquito
control strategies.
Methods
Mosquitoes and xenobiotics
A laboratory strain of the dengue vector Ae. aegypti
(Bora-Bora strain), susceptible to insecticides was reared
in standard insectary conditions (26°C, 8 h/16 h light/
dark period) and used for all experiments. Larvae were
reared in tap water with controlled amount of larval food
(ground hay pellets) for 4 days (3rd instar) before exposure
for 48 h to 3 chemical insecticides and 3 pollutants
belonging to various chemical classes: the pyrethroid
insecticide permethrin (Chem Service, USA), the neonic-
otinoid insecticide imidacloprid (Sigma Aldrich, USA),
the carbamate insecticide propoxur (Sigma Aldrich,
USA), the herbicide atrazine (Cluzeau, France), the poly-
cyclic aromatic hydrocarbon (PAH) fluoranthene
(Aldrich, France) and the heavy metal copper (obtained
from CuSO4, Prolabo, France). Atrazine is an herbicide
heavily used worldwide and is likely to be found in mos-
quito breeding sites near cultivated areas (e.g. field drain-
pipes) [30,55]. Similarly, copper is the major component
of Bordeaux mixture and is widely used to control fungus
on grapes and other berries [56]. Finally, fluoranthene is
one of the most ubiquitous PAH and is found at high con-
centrations in road sediments [57]. Elevated doses of flu-
oranthene are likely to be found in urban mosquito
breeding sites such as road trenches [58] or in oil spillage
areas [4].
Samples preparation
Exposures to all xenobiotics were performed in triplicate
with larvae from different egg batches (3 biological repli-
cates per treatment). One hundred larvae were exposed
to each xenobiotic in 200 ml tap water containing 50 mg
of larval food. Control larvae were obtained simultane-
ously in similar conditions without xenobiotics. Doses of
xenobiotics used for larval exposure were chosen accord-
ing to the doses likely to be found in highly polluted mos-
quito breeding sites (INERIS, http://www.ineris.fr).
Preliminary experiments revealed that fluoranthene,
atrazine or copper did not show any toxicity on mosquito
larvae even at higher concentrations than those used in
the present study. For insecticides, we chose a concentra-
tion resulting in less than 15% larval mortality after 48 h
exposure. This low mortality threshold was chosen inlands by anthropogenic chemicals and the role of pheno-
typic plasticity in driving selection mechanisms [54] raise
the question of the long-term impact of pollutants on the
order to minimize the effect of the artificial selection of
particular genotypes more tolerant to the insecticide dur-
ing exposure. Doses of xenobiotics used for exposures

David et al. BMC Genomics 2010, 11:216
http://www.biomedcentral.com/1471-2164/11/216
Page 9 of 12
were 1.5 μg/L permethrin, 40 μg/L imidacloprid, 500 μg/L
propoxur, 25 μg/L fluoranthene, 10 μg/L atrazine and 2
mg/L CuSO4. After 48 h, larvae were collected, rinsed
twice in tap water and immediately used for RNA extrac-
tions.
Preparation of double stranded cDNA tag libraries
For each biological replicate, total RNA was extracted
from 30 fresh larvae using the PicoPure™ RNA isolation
kit (Arcturus Bioscience, Mountain View, USA) accord-
ing to manufacturer's instructions. Total RNA quality and
quantity were controlled on an Agilent 2100 Bioanalyzer
(Agilent, USA). Total RNAs were then diluted to 750 ng/
μL in nuclease-free water. For each treatment, total RNAs
from the 3 biological replicates were then pooled
together in equal proportions. Double-stranded cDNA
tag libraries (Additional file 5: Suppl. Figure 4) were pre-
pared by Illumina Corporation. Two μg total RNA were
used to isolate mRNAs by using magnetic oligo(dT)
beads before cDNA synthesis using superscript II (Invit-
rogen) at 42°C for 1 h. Second strand cDNAs were then
synthesized and mRNAs were removed. Double stranded
cDNAs were cleaved at DpnII restriction sites (5'-
GATC-3') and fragments attached to the oligo(dT) beads
on their 3' end were purified. Gene expression (GEX)
adapters 1 were ligated to the DpnII cleavage sites using
T4 DNA ligase (Invitrogen). Double stranded cDNAs
containing both GEX adaptors 1 and oligo(dT) beads
were then digested with MmeI for 1.5 h at 37°C to gener-
ate 20 bp double stranded cDNA tags. These tags were
purified before ligating GEX adapters 2 at the MmeI
cleavage site using T4 DNA ligase. The adapter-ligated
cDNA tag library was then enriched by PCR with two
primers annealing to the end of GeX adapters and Phu-
sion DNA polymerase (Finnzymes Oy). PCR cycles were
30 s at 98°C followed by 15 cycles of 10 s at 98°C, 30 s at
60°C, 15 s at 72°C and a final elongation step of 10 min at
72°C. Sequences of primers used for library preparation
are available at http://illumina.com. Enriched cDNA tag
library was then gel-purified before quality control analy-
sis on an Agilent 2100 Bioanalyzer.
Sequencing and mapping of cDNA tags to mosquito
genome
Each cDNA tag library was sequenced as 20-mers on a
genome analyzer I (illumina Corporation). Each cDNA
tag library was sequenced on a separated flow cell lane.
Sequenced cDNA tags were then filtered from back-
ground noise according to their total number of reads
across all conditions. Only cDNA tags represented by
more than 20 reads were kept for further analysis. Back-
based on the short sequence mapping algorithm 'agrep'
[59]. TagMatcher allows matching tags to a reference
genome with errors and multiple matching loci (available
on request to
eric.coissac@inrialpes.fr
). After mapping to Ae. aegypti genome, only tags without
ambiguous nucleotides and mapped without mismatch at
a unique genomic location were kept for clustering and
differential transcription analysis. To avoid possible bias
due to incomplete 3' UTR annotation and because most
cDNA tags were expected on the 3' side of genes (see
Additional file 5: Suppl. Figure 4), cDNA tags were con-
sidered to be 'within' a gene if located between the 5'
boundary of a gene and its 3' boundary extended by 300
bp.
Clustering and differential transcription analysis
In order to collect transcription data from distinct tags
matching to a unique transcript or a unique genomic loci
without a priori knowledge of genome annotation, we
clustered tags previously mapped to Ae. aegypti genome.
Two distinct tags were assigned to a single cluster if i)
tags were found on the same DNA strand and genomic
supercontig, ii) tags were separated by less than 500 bp
and iii) the total number of reads across all conditions
was higher for the tag located downstream (3' side) than
for the tag located upstream (5' side). The later condition
was adopted in order to take in account the effect of par-
tial DpnII digestion of cDNAs during cDNA library prep-
aration, leading to multiple tags located on a single
transcript with decreasing number of reads toward the 5'
direction (see Additional file 5: Suppl. Figure 4).
Differential analysis of transcription levels in mosquito
larvae exposed to each xenobiotic was performed at the
gene level for cDNA tags mapped within predicted genes
(i.e. gathering all tags mapped within each gene) and at
the cluster level for cDNA tags not mapped within pre-
dicted genes (i.e. gathering all tags mapped within each
cluster). Transcription ratios (TR) were calculated by
dividing the number of reads per million (RPM) in xeno-
biotic-exposed larvae by the number of RPM in control
larvae following the formula: TR = [(RPMtreated + x)/
(RPMcontrols + x)], where x is a pseudocount equal to 0.2
(approximately 1 read per million per condition). Then,
the probability of each gene to be differentially tran-
scribed more than 2-fold in either direction between
treated and controls was computed for each condition
from raw read counts, taking into account library size.
This computation was performed using Fisher's noncen-
tral hypergeometric distribution, which has the advan-
tage over standard hypergeometric law to allowground-filtered cDNA tags were then mapped to the Ae.
aegypti genome assembly (AaegL 1.1 annotation) using
TagMatcher, a software developed in our laboratory and
computation of Pvalue for a ratio different of one [60].
Holm correction was then applied to multiple test proce-
dure. Genes/clusters were considered differentially tran-

David et al. BMC Genomics 2010, 11:216
http://www.biomedcentral.com/1471-2164/11/216
Page 10 of 12
scribed between xenobiotic-exposed larvae and controls
if Pvalue < 10-3.
Differential effect of xenobiotics on mosquito larvae
transcriptome
To compare the global effect of each xenobiotic on Ae.
aegypti larvae transcriptome, a principal component
analysis (PCA) based on Log10 TRs was performed on the
453 genes and 225 clusters not mapped within genes
showing significant differential transcription following
exposure to at least one xenobiotic. Representation of
observations (genes and clusters) and conditions (xenobi-
otics used for exposure) on PCA axis was optimized by
applying a Varimax rotation on the 5 axis best represent-
ing the variance [61]. A comparative analysis of gene
functions differentially transcribed was performed on the
453 genes showing significant differential transcription
following exposure to at least one xenobiotic. Genes were
classified in 9 different categories: enzymes, kinases,
transport, DNA interaction, cuticle, cytoskeleton, ribo-
somes, others and hypothetical proteins. For each treat-
ment, percentages of genes significantly over- and under-
transcribed were compared. To investigate the role of
enzymes in the response of mosquito larvae to xenobiot-
ics, a hierarchical clustering analysis based on TRs was
performed on the 115 enzymes showing a significant dif-
ferential transcription. Clustering analysis was performed
by loading Log10 transcription ratios into TM4 Multi
experiment Viewer (MeV) software [62]. Gene and condi-
tion trees were calculated using Pearson's uncentered dis-
tance metric and complete linkage method with
optimization of genes order [63,64].
Real-time quantitative RT-PCR validation
Transcription profiles of 14 genes were validated by
reverse transcription followed by real-time quantitative
PCR on same RNA samples used for cDNA library prepa-
ration. Four μg total RNAs were treated with DNAse I
(Invitrogen) and used for cDNA synthesis with super-
script III (Invitrogen) and oligo-dT20 primer according to
manufacturer's instructions. Resulting cDNAs were
diluted 100 times for PCR reactions. Real-time quantita-
tive PCR reactions of 25 μL were performed in triplicate
on an iQ5 system (BioRad) using iQ SYBR Green super-
mix (BioRad), 0.3 μM of each primer and 5 μL of diluted
cDNAs according to manufacturer's instructions. Data
analysis was performed according to the ΔΔCT method
taking into account PCR efficiency [65] and using the two
genes encoding the ribosomal protein L8 (GenBank
accession no. DQ440262) and the ribosomal protein S7
(Genbank accession no. EAT38624.1) for normalisation.
Data deposition
Detailed transcription data for the 6850 genes detected in
the present study are presented in the Additional file 6
(supplementary Table 2).
All next-generation sequencing data and cDNA library
informations associated to the present study have been
deposited at the EMBL-EBI European Read Archive
(ERA) under accession number ERA000115. Experiment
metadata are freely accessible at ftp://ftp.era-
xml.ebi.ac.uk/meta/xml/ and sequence data are freely
accessible at ftp://ftp.era-xml.ebi.ac.uk/vol1/ERA000/
ERA000115/. Expression data from the 453 genes found
differentially transcribed after xenobiotic exposure are
also accessible at http://funcgen.vectorbase.org/Expres-
sionData/.
All gene accession numbers mentioned in the present
manuscript are compatible with Ensembl, NCBI-Gen-
Bank and Vectorbase http://aaegypti.vectorbase.org
genome databases.
Additional material
Additional file 1 Supplementary figure 1. This figure represents the dis-
tribution of the number of reads across distinct genes (6850 genes), clusters
not mapped within predicted genes (4868 clusters), all mapped clusters
(13118 clusters) and all mapped tags (15253 tags). Genes, clusters and tags
are ranked in ascending order according to their total number of reads
across all conditions.
Additional file 2 Supplementary table 1. This table contains all tran-
scription data for the 453 genes found differentially transcribed in Aedes
aegypti larvae exposed to xenobiotics. Genes are arranged in nine different
functional categories: enzymes; kinases; transport; DNA interaction; cuticle;
cytoskeleton; ribosomes; others and unknown hypothetical proteins. For
each gene, accession number and gene name or annotation are indicated.
The number of reads per million (RPM) across all conditions is indicated as
an average transcription level. Log10 transcription ratios (exposed to xeno-
biotic/control) are indicated for each xenobiotic relative to control. Tran-
scription ratios with a significant Fisher's test Pvalue < 0.001 are shown in
bold.
Additional file 3 Supplementary figure 2. This figure shows the valida-
tion of transcription ratios obtained from Digital Gene Expression Tag Profil-
ing (DGETP) by real-time quantitative RT-PCR. Validation was performed on
14 genes found significantly over-transcribed by DGETP in at least one con-
dition. For each gene, transcription ratios from both techniques across all
conditions are represented. Black dots represent conditions showing a sig-
nificant over-transcription in DGETP. Accession numbers and annotations of
gene analyzed were: AAEL001626 (zinc/iron transporter); AAEL001981 (ser-
ine/threonine kinase); AAEL002110 (cuticular protein); AAEL004748 (pupal
cuticular protein); AAEL004829 (NADH dehydrogenase); AAEL005416 (oxi-
dase/peroxidase); AAEL005696 (cytochrome P450 CYP325X2); AAEL005929
(ATP-binding cassette transporter); AAEL010500 (glutathione S-transferase
GSTX2); AAEL011008 (lipase); AAEL012636 (cytochrome b5); AAEL013514
(pupale cuticle protein); AAEL009127 (cytochrome P450 CYP6M11);
AAEL001807 (cytochrome P450 CYP9M9).
Additional file 4 Supplementary figure 3. This figure represents the
results of the principal component analysis of the effect of xenobiotics on
mosquito larvae transcriptome. Analysis was based on log10 transcription
ratios of all genes and clusters not mapped within genes showing a signifi-
cant differential transcription in at least one treatment. Both xenobiotic For each treatment, results were expressed as mean tran-
scription ratios (± SE) between xenobiotic-exposed larvae
and control larvae.
treatments (black dots) and genes or clusters (grey crosses) are represented
using the 3 axis best representing the variance. Biplot A: axis 1 and 2 (81.5%
of variance). Biplot B: axis 1 and 3 (69.7% of variance).

David et al. BMC Genomics 2010, 11:216
http://www.biomedcentral.com/1471-2164/11/216
Page 11 of 12
Authors' contributions
JPD conceived and coordinated the study, participated in sample preparation
and data analysis and wrote the manuscript. EC and CM performed bioinfor-
matics and statistical analysis and help to draft the manuscript. RP and MAR
performed qRT-PCR experiments, contributed to sample preparation, data
analysis and help to draft the manuscript. ACP contributed to data analysis and
helped to draft the manuscript. SR contributed to study design, sample prepa-
ration, data analysis and helped writing the manuscript. All authors read and
approved the final manuscript.
Acknowledgements
The present research project was funded by the French National Research
Agency (ANR project 07SEST014 MOSQUITO-ENV). We are grateful to J. Patou-
raux and S. Veyrenc for technical help. We thank Dr. B. MacCallum from Vector-
base and Dr. C. Hunter and Dr. V. Zalunin from EMBL-EBI for help with sequence
data deposition. We are grateful to Dr. J. Vontas, Dr. P. Taberlet, Dr. H. Ranson, Pr.
P. Ravanel and anonymous reviewers for useful comments on the manuscript.
Author Details
Laboratoire d'Ecologie Alpine (LECA, UMR 5553 CNRS - Université Grenoble),
France
References
1. Stegeman JJ, Hahn ME: Biochemistry and molecular biology of
monooxygenase: current perspective on forms, functions, and
regulation of cytochrome P450 in aquatic species. In Aquatic toxicology;
Molecular, Biochemical and Cellular Perspectives Edited by: Malins DC,
Ostrander GK. Boca Raton: Lewis Publisher, CRC Press; 1994:87-206.
2. Lounibos LP: Invasions by insect vectors of human disease. Annu Rev
Entomol 2002, 47:233-266.
3. Hemingway J, Field L, Vontas J: An overview of insecticide resistance.
Science 2002, 298(5591):96-97.
4. Djouaka RF, Bakare AA, Coulibaly ON, Akogbeto MC, Ranson H,
Hemingway J, Strode C: Expression of the cytochrome P450s, CYP6P3
and CYP6M2 are significantly elevated in multiple pyrethroid resistant
populations of Anopheles gambiae s.s. from Southern Benin and
Nigeria. BMC Genomics 2008, 9:538.
5. Liu NN, Liu HQ, Zhu F, Zhang L: Differential expression of genes in
pyrethroid resistant and susceptible mosquitoes, Culex
quinquefasciatus (S.). Gene 2007, 394(1-2):61-68.
6. Marcombe S, Poupardin R, Darriet F, Reynaud S, Bonnet J, Strode C,
Brengues C, Yebakima A, Ranson H, Corbel V, et al.: Exploring the
molecular basis of insecticide resistance in the dengue vector Aedes
aegypti: a case study in Martinique Island (French West Indies). BMC
Genomics 2009, 10:494.
7. Muller P, Donnelly MJ, Ranson H: Transcription profiling of a recently
colonised pyrethroid resistant Anopheles gambiae strain from Ghana.
resistance in malaria vectors. Proc Natl Acad Sci USA 2005,
102(11):4080-4084.
9. Strode C, Wondji CS, David JP, Hawkes NJ, Lumjuan N, Nelson DR, Drane
DR, Karunaratne S, Hemingway J, Black WC, et al.: Genomic analysis of
detoxification genes in the mosquito Aedes aegypti. Insect Biochem Mol
Biol 2008, 38(1):113-123.
10. Vontas J, Blass C, Koutsos AC, David JP, Kafatos FC, Louis C, Hemingway J,
Christophides GK, Ranson H: Gene expression in insecticide resistant
and susceptible Anopheles gambiae strains constitutively or after
insecticide exposure. Insect Mol Biol 2005, 14(5):509-521.
11. Poupardin R, Reynaud S, Strode C, Ranson H, Vontas J, David JP: Cross-
induction of detoxification genes by environmental xenobiotics and
insecticides in the mosquito Aedes aegypti: Impact on larval tolerance
to chemical insecticides. Insect Biochem Mol Biol 2008, 38(5):540-551.
12. Riaz MA, Poupardin R, Reynaud S, Strode C, Ranson H, David JP: Impact of
glyphosate and benzo[a]pyrene on the tolerance of mosquito larvae to
chemical insecticides. Role of detoxification genes in response to
xenobiotics. Aquat Toxicol 2009, 93(1):61-69.
13. Sasabe M, Wen ZM, Berenbaum MR, Schuler MA: Molecular analysis of
CYP321A1, a novel cytochrome P450 involved in metabolism of plant
allelochernicals (furanocoumarins) and insecticides (cypermethrin) in
Helicoverpa zea. Gene 2004, 338(2):163-175.
14. Green CD, Simons JF, Taillon BE, Lewin DA: Open systems: panoramic
views of gene expression. J Immunol Methods 2001, 250(1-2):67-79.
15. Hanriot L, Keime C, Gay N, Faure C, Dossat C, Wincker P, Scote-Blachon C,
Peyron C, Gandrillon O: A combination of LongSAGE with Solexa
sequencing is well suited to explore the depth and the complexity of
transcriptome. BMC Genomics 2008, 9:418.
16. Velculescu VE, Zhang L, Vogelstein B, Kinzler KW: Serial analysis of gene-
expression. Science 1995, 270(5235):484-487.
17. Saha S, Sparks AB, Rago C, Akmaev V, Wang CJ, Vogelstein B, Kinzler KW,
Velculescu VE: Using the transcriptome to annotate the genome. Nat
Biotechnol 2002, 20(5):508-512.
18. Brenner S, Johnson M, Bridgham J, Golda G, Lloyd DH, Johnson D, Luo SJ,
McCurdy S, Foy M, Ewan M, et al.: Gene expression analysis by massively
parallel signature sequencing (MPSS) on microbead arrays. Nat
Biotechnol 2000, 18(6):630-634.
19. Nielsen KL, Hogh AL, Emmersen J: DeepSAGE - digital transcriptomics
with high sensitivity, simple experimental protocol and multiplexing of
samples. Nucleic Acids Res 2006, 34(19):e133.
20. Eveland AL, McCarty DR, Koch KE: Transcript profiling by 3'-untranslated
region sequencing resolves expression of gene families. Plant Physiol
2008, 146(1):32-44.
21. Nene V, Wortman JR, Lawson D, Haas B, Kodira C, Tu ZJ, Loftus B, Xi ZY,
Megy K, Grabherr M, et al.: Genome sequence of Aedes aegypti, a major
arbovirus vector. Science 2007, 316(5832):1718-1723.
22. Sackton TB, Clark AG: Comparative profiling of the transcriptional
response to infection in two species of Drosophila by short-read cDNA
sequencing. BMC Genomics 2009, 10:259.
23. Kapranov P, Willingham AT, Gingeras TR: Genome-wide transcription
and the implications for genomic organization. Nat Rev Genet 2007,
8(6):413-423.
24. Mattick JS: The Genetic Signatures of Noncoding RNAs. PLoS Genet
2009, 5(4):e1000459.
25. Roche H, Buet A, Ramade F: Accumulation of lipophilic
microcontaminants and biochemical responses in eels from the
Camargue Biosphere Reserve. Ecotoxicology 2002, 11(3):155-164.
26. Sandermann H: Differential lipid affinity of xenobiotics and natural
compounds. FEBS Lett 2003, 554(1-2):165-168.
27. Tomizawa M, Casida JE: Neonicotinoid insecticide toxicology:
Mechanisms of selective action. Annu Rev Pharmacol Toxicol 2005,
45:247-268.
28. Raymond-Delpech V, Matsuda K, Sattelle BM, Rauh JJ, Sattelle DB: Ion
channels: molecular targets of neuroactive insecticides. Invert Neurosci
2005, 5(3-4):119-133.
29. Sherman JD: Organophosphates pesticides - Neurological and
respiratory toxicity. Toxicol Ind Health 1995, 11(1):33-39.
30. Boyer S, Serandour J, Lemperiere G, Raveton M, Ravanel P: Do herbicide
Additional file 5 Supplementary figure 4. This figure illustrates the prep-
aration of the double stranded cDNA tag library. Messenger RNAs are iso-
lated by using magnetic oligo(dT) beads before cDNA synthesis. Double
stranded cDNAs are synthesized using DNA polymerase I and clived at
every DpnII restriction sites. Gene expression (GEX) adapters 1 containing a
MmeI recognition site at its 3' side are then ligated to the DpnII clivage sites.
Double stranded cDNA fragments containing both GEX adaptor 1 and
oligo(dT) beads were then digested with MmeI to generate double
stranded cDNA tags. These tags were purified and ligated with GEX adapt-
ers 2 at the MmeI cleavage site before enrichment by PCR.
Additional file 6 Supplementary table 2. This table describes transcrip-
tion data for the 6850 genes detected in the study. For each gene, acces-
sion number and annotation are indicated. For each gene and each
condition, reads counts, reads per millions (RPM), transcription ratios (Log10
TR) and Fisher's test Pvalues are indicated.
Received: 6 January 2010 Accepted: 31 March 2010
Published: 31 March 2010
This article is available from: http://www.biomedcentral.com/1471-2164/11/216© 2010 David et al; licensee BioMed Central Ltd. is an Open Access article distributed under th terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.BMC Ge omics 2010, 11:216BMC Genomics 2007, 8:36.
8. David JP, Strode C, Vontas J, Nikou D, Vaughan A, Pignatelli PM, Louis C,
Hemingway J, Ranson H: The Anopheles gambiae detoxification chip: A
highly specific microarray to study metabolic-based insecticide
treatments reduce the sensitivity of mosquito larvae to insecticides?
Chemosphere 2006, 65(4):721-724.

David et al. BMC Genomics 2010, 11:216
http://www.biomedcentral.com/1471-2164/11/216
Page 12 of 12
31. Reynaud S, Deschaux P: The effects of polycyclic aromatic hydrocarbons
on the immune system of fish: A review. Aquat Toxicol 2006,
77(2):229-238.
32. Chiu TL, Wen ZM, Rupasinghe SG, Schuler MA: Comparative molecular
modeling of Anopheles gambiae CYP6Z1, a mosquito P450 capable of
metabolizing DDT. Proc Natl Acad Sci USA 2008, 105(26):8855-8860.
33. Daborn PJ, Yen JL, Bogwitz MR, Le Goff G, Feil E, Jeffers S, Tijet N, Perry T,
Heckel D, Batterham P, et al.: A single P450 allele associated with
insecticide resistance in Drosophila. Science 2002,
297(5590):2253-2256.
34. Despres L, David JP, Gallet C: The evolutionary ecology of insect
resistance to plant chemicals. Trends in Ecology & Evolution 2007,
22(6):298-307.
35. Vontas J, David JP, Nikou D, Hemingway J, Christophides GK, Louis C,
Ranson H: Transcriptional analysis of insecticide resistance in Anopheles
stephensi using cross-species microarray hybridization. Insect Mol Biol
2007, 16(3):315-324.
36. Zhang J, Goyer C, Pelletier Y: Environmental stresses induce the
expression of putative glycine-rich insect cuticular protein genes in
adult Leptinotarsa decemlineata (Say). Insect Mol Biol 2008,
17(3):209-216.
37. Selye H: The stress of life, Revised Edition. In Circuits and systems in stress:
preclinical studies. Depression and Anxiety Volume 15. Edited by: Vermetten
E, Bremner D. New York: McGraw-Hill; 1976:126-147.
38. Sanders BM, Martin LS: Stress protein as biomarkers of contaminant
exposure in archived environmental-samples. Sci Total Environ 1993,
140:459-470.
39. Lee SM, Lee SB, Park CH, Choi J: Expression of heat shock protein and
hemoglobin genes in Chironomus tentans (Diptera, chironomidae)
larvae exposed to various environmental pollutants: A potential
biomarker of freshwater monitoring. Chemosphere 2006,
65(6):1074-1081.
40. Balamurugan K, Hua HQ, Georgiev O, Schaffner W: Mercury and
cadmium trigger expression of the copper importer Ctr1B, which
enables Drosophila to thrive on heavy metal-loaded food. Biol Chem
2009, 390(2):109-113.
41. Vache C, Camares O, De Graeve F, Dastugue B, Meiniel A, Vaury C, Pellier S,
Leoz-Garziandia E, Bamdad M: Drosophila melanogaster P-glycoprotein:
A membrane detoxification system toward polycyclic aromatic
hydrocarbon pollutants. Environ Toxicol Chem 2006, 25(2):572-580.
42. Palmfeldt J, Vang S, Stenbroen V, Pedersen CB, Christensen JH, Bross P,
Gregersen N: Mitochondrial proteomics on human fibroblasts for
identification of metabolic imbalance and cellular stress. Proteome
Science 2009, 7:20.
43. Holth TF, Nourizadeh-Lillabadi R, Blaesbjerg M, Grung M, Holbech H,
Petersen GI, Alestrom P, Hylland K: Differential gene expression and
biomarkers in zebrafish (Danio rerio) following exposure to produced
water components. Aquat Toxicol 2008, 90(4):277-291.
44. Owen J, Hedley BA, Svendsen C, Wren J, Jonker MJ, Hankard PK, Lister LJ,
Sturzenbaum SR, Morgan AJ, Spurgeon DJ, et al.: Transcriptome profiling
of developmental and xenobiotic responses in a keystone soil animal,
the oligochaete annelid Lumbricus rubellus. BMC Genomics 2008, 9:266.
45. Reynaud S, Duchiron C, Deschaux P: 3-Methylcholanthrene induces
lymphocyte and phagocyte apoptosis in common carp (Cyprinus
carpio L) in vitro. Aquat Toxicol 2004, 66(3):307-318.
46. Reynaud S, Raveton M, Ravanel P: Interactions between immune and
biotransformation systems in fish: A review. Aquat Toxicol 2008,
87(3):139-145.
47. Knauert S, Dawo U, Hollender J, Hommen U, Knauer K: Effects of
photosystem II inhibitors and their mixture on
freshwaterphytoplancton succession in outdoor mesocosms. Environ
Toxicol Chem 2009, 28(4):836-845.
48. Knauert S, Escher B, Singer H, Hollender J, Knauer K: Mixture toxicity of
three photosystem II inhibitors (atrazine, isoproturon, and diuron)
toward photosynthesis of freshwater phytoplankton studied in
outdoor mesocosms. Environ Sci Technol 2008, 42(17):6424-6430.
49. Harris ED: Cellular copper transport and metabolism. Annu Rev Nutr
2000, 20:291-310.
51. Letelier ME, Martinez M, Gonzalez-Lira V, Faundez M, Aracena-Parks P:
Inhibition of cytosolic glutathione S-transferase activity from rat liver
by copper. Chem Biol Interact 2006, 164(1-2):39-48.
52. Posthuma L, Eijsacker HJP, Koelmans AA, Vijver MG: Ecological effects of
diffuse mixed pollution are site-specific and require higher-tier risk
assessment to improve site management decisions: A discussion
paper. Sci Total Environ 2008, 406(3):503-517.
53. Wurff AW Van der, Kools SA, Boivin ME, Brink PJ Van den, Van Megen HH,
Riksen JA, Doroszuk A, Kammenga JE: Type of disturbance and
ecological history determine structural stability. Ecol Appl 2007,
17(1):190-202.
54. Ghalambor CK, McKay JK, Carroll SP, Reznick DN: Adaptive versus non-
adaptive phenotypic plasticity and the potential for contemporary
adaptation in new environments. Functional Ecology 2007,
21(3):394-407.
55. Gruessner B, Watzin MC: Response of aquatic communities from a
Vermont stream to environmentally realistic atrazine exposure in
laboratory microcosms. Symposium on Methods for Monitoring Effects of
Pesticides, at the 15th Annual Meeting of the Society-of-Environmental-
Toxicology-and-Chemistry: Oct 30-Nov 03 1994; Denver, Co 1994:410-419.
56. Komarek M, Vanek A, Chrastny V, Szakova J, Kubova K, Drahota P, Balik J:
Retention of copper originating from different fungicides in
contrasting soil types. J Hazard Mater 2009, 166(2-3):1395-1402.
57. Durand C, Ruban V, Ambles A, Clozel B, Achard L: Characterisation of
road sediments near Bordeaux with emphasis on phosphorus. J
Environ Monit 2003, 5(3):463-467.
58. Zehetner F, Rosenfellner U, Mentler A, Gerzabek MH: Distribution of Road
Salt Residues, Heavy Metals and Polycyclic Aromatic Hydrocarbons
across a Highway-Forest Interface. Water Air and Soil Pollution 2009,
198(1-4):125-132.
59. Wu S, Manber U: Agrep - a fast approximate pattern-matching tool.
Proceedings USENIX Winter 1992 Technical Conference: 1992; San Francisco,
CA 1992:153-162.
60. Fisher RA: The logic of inductive inference. Journal of the Royal Statistical
Society Series A 1935, 98:39-54.
61. Kaiser HF: The varimax criterion for analytic rotation in factor analysis.
Psychometrika 1958, 23(3):187-200.
62. Saeed AI, Sharov V, White J, Li J, Liang W, Bhagabati N, Braisted J, Klapa M,
Currier T, Thiagarajan M, et al.: TM4: a free, open-source system for
microarray data management and analysis. Biotechniques 2003,
34(2):374-378.
63. D'haeseleer P: How does gene expression clustering work? Nat
Biotechnol 2005, 23(12):1499-1501.
64. Eisen MB, Spellman PT, Brown PO, Botstein D: Cluster analysis and
display of genome-wide expression patterns. Proc Natl Acad Sci USA
1998, 95:14863-14868.
65. Pfaffl MW: A new mathematical model for relative quantification in real-
time RT-PCR. Nucleic Acids Res 2001, 29(9):e45.
doi: 10.1186/1471-2164-11-216
Cite this article as: David et al., Transcriptome response to pollutants and
insecticides in the dengue vector Aedes aegypti using next-generation
sequencing technology BMC Genomics 2010, 11:21650. Krupanidhi S, Sreekumar A, Sanjeevi CB: Copper & biological health.
Indian J Med Res 2008, 128(4):448-461.