Transcriptional responses of Anopheles gambiae s . s mosquito larvae to chronic exposure of cadmium heavy metal

Background: Anopheles gambiae larvae traditionally thrive in non-polluted environments. We previously documented the presence of the larvae in heavy metal polluted urban aquatic environments and the associated biological cost. The goal of this study was to unravel the molecular dynamics involved in the adaptation of the mosquitoes to the heavy metals. Methods: Total RNA was extracted from third instar larvae of both cadmium treated populations and untreated control populations. The RNA concentrations were normalized and complementary DNAs were prepared. Then annealing control primer (ACP) technology was applied to establish transcriptional responses in An. gambiae larvae following several generational (n=90) chronic exposures to cadmium. Differentially expressed genes were determined by their differential banding patterns on an agarose gel. Gel extraction and purification was then carried out on the DEGs and these were later cloned and sequenced to establish the specific transcripts.    Results: We identified 14 differentially expressed transcripts in response to the cadmium exposure in the larvae. Most (11) of the transcripts were up-regulated in response to the cadmium exposure and were putatively functionally associated with metabolism, transport and protein synthesis processes. The transcripts included ATP-binding cassette transporter, eupolytin, ribosomal RNA, translation initiation factor, THO complex, lysosomal alpha-mannosidase, sodium-independent sulfate anion transporter and myotubularin related protein 2. The down-regulated transcripts were functionally associated with signal transduction and proteolytic activity and included Protein G12, adenylate cyclase and endoplasmic reticulum metallopeptidase. Conclusions: Our findings shed light on pathways functionally associated with the adaptation to heavy metals that can be targeted in integrated vector control programs, and potential An. gambiae larvae biomarkers for assessment of environmental stress or contamination.


Introduction
Heavy metal pollution has become a global environmental problem and severely threatens biological diversity and human health. Our studies on adaptation to heavy metals have documented presence of the mosquitoes in polluted habitats (Mireji et al., 2008) with growing evidence that this adaptation comes at a biological cost to the mosquito (Mireji et al., 2010b). Similar biological costs to adaptations have also been observed elsewhere in Culex pipiens L responses to cadmium, copper, lead and mercury (El-Sheikh et al., 2010). To date, molecular dynamics underpinning heavy metal tolerance in insects have been tied to transcripts and genes associated functionally with immunity (Sorvari et al., 2007) and defense and repair mechanisms such as glutathione transferases and heat shock proteins (Liao & Freedman, 1998;Kim et al., 2000;Stohs et al., 2001). We have previously putatively implicated metallothioneins, alpha-tubulin and cytochrome p450 genes associated with defense, repair and pyrethroid metabolism mechanisms in insects with heavy metal tolerance, using single gene assessment approaches with Anopheles gambiae mosquito larvae (Mireji et al., 2010b;Mireji et al., 2006;Musasia et al., 2013). Here, we have emulated ab initio relatively higher throughput annealing control primer (ACP) transcriptional profiling, to identify: 1) Pathways functionally associated with heavy metal adaptation observed in the field and their associated biological costs (Mireji et al., 2008;Mireji et al., 2010b); and 2) Potential An. gambiae larvae biomarkers that can be applied for assessment of environmental stress or contamination.  (Mireji et al., 2010a). Cadmium metal tolerant strains and control (untreated) strains of the mosquito were raised separately and in triplicates. All subsequent generations of the mosquito were subjected to chronic exposures of cadmium metal as described in Mireji et al., (2010a). Standard Operating Procedure (SOP) for the rearing of Anopheles mosquitoes was followed for colony maintenance (Ford & Green, 1972). Cadmium used in our study was applied as Cadmium Chloride (CdCl 2 ) 99.99% pure (Fisher Scientific LLC, Fair Lawn, NJ, U.S.A).

RNA isolation
Total RNA was extracted from the third instar larvae of experimental and control An. gambiae populations using Trizol (Invitrogen). Quantification of the extracted RNA was done using the microspectrophotometer Genequant pro (Amersham Pharmacia Ltd., Bucks, UK) (Table 1) and a few samples were quantified using the NanoDrop ® (Therm Scientific) ( Table 2). In addition, DNaseI digestion was carried out to remove any residual DNA that could present in the extracted RNA. Total RNA that was isolated was stored at -80°C.

GeneFishing ™ Reverse Transcription
The total RNA extracted from experimental and control An. gambiae populations were normalized to same concentrations and directly used for the synthesis of first strand complementary DNA (cDNA) using reverse transcriptase (Hwang et al., 2003).
ACP based-GeneFishing ™ PCR Annealing control primer based PCR using the GeneFishing™ DEG kit from Seegene, Seoul, South Korea (Kim et al., 2004), was used to determine differentially expressed genes in the heavy metal treated group and the control population.
Synthesis of the second strand cDNA and PCR was carried out in a single tube. The second strand was synthesized in one cycle of first stage PCR at 50°C, in a final reaction volume of 20µl. The components in the reaction tubes included 3-5µl of diluted first

Amendments from Version 1
In the previous version (check in the materials and methods section-RNA isolation) we had only stated that total RNA that had been extracted was quantified using the micro-spectrophotometer Genequant pro because this was the main equipment that we used for quantification of RNA and we had not recorded the outcome of the quantification.
In the new version, we have included quantification that had also been done for a few samples using the NanoDrop ® (Thermo Scientific). The quantification outcome which demonstrates the RNA integrity, has now been recorded in Table 1 for RNA quantification using the micro-spectrophotometer Genequant pro, and in Table 2 for RNA quantification using the NanoDrop ® (Thermo Scientific).
RNA is susceptible to degradation due to the presence of RNases in the environment. High quality RNA is required for downstream applications and therefore RNA must be quantified to ensure its integrity. To estimate RNA purity, the ratio of the Absorbance contributed by the RNA to the Absorbance of the contaminants is calculated. The acceptable ratios for purity or typical requirements for A260/A280 ratios are 1.8-2.2.
The extracted RNAs that were considered for the downstream application in Gene fishing Technology were those that met this criterion of purity as stated above.
The added information was in response to the question raised by: Shüné V. Oliver. The other question that had previously been raised by David Essumang, about the generations, I responded to his comment and the information he had requested was already present in the manuscript. strand cDNA, 1µl 10Mm dT-ACP2 reverse primer (5'-CTGTG AATGCTGCGACTACGATIIIII(T) 15 -3'), 10µl 2x master mix (Seegene, Seoul, South Korea) and 1µl 10µM arbitrary ACP (forward primer).

REVISED
PCR procedures for the synthesis of the second strand were completed in one cycle, at 94°C for 1 min then 50°C for 3min and 72°C for 1 min.
The second stage of the PCR protocol consisted of 40 cycles at 94°C for 40s, 65°C for 40s, 72°C for 40s and the final extension for 10 min at 72°C. 2% agarose gel electrophoresis with ethidium bromide staining was used for separation of the PCR products.

Gel extraction
Differentially expressed bands in the control and cadmium exposed population subjected to the same primer set were excised from the agarose gel using a scapel under Ultra Violet illumination. The gel slices were then purified using the QIAquick ® gel extraction kit (QIAGEN, Inc., Valencia, CA), following the instructions from the manufacturer.

Cloning
Gel-purified PCR products were directly cloned into a pGEMT Easy vector (Invitrogen, Carlsbad, CA, USA), using JM109 competent cells. Colonies were grown at 37°C for 18 hours on Luria broth agar plates, containing ampicillin, X-gal and IPTG for blue/white colony screening. Cloned plasmids were then purified using the GeneJET ™ Miniprep kit (Fermentus, Thermo Fisher Scientific Inc.), as per the instructions from the manufacturer.

Sequencing
Sequencing was done with ABI PRISM ® 3100 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA) using M13 primers. The sequences were edited using VecScreen and BioEdit software. Edited sequences were analyzed by searching for similarities in VectorBase against the Anopheles gambiae PEST strain transcripts sequences, AgamP4.6 geneset using the BLASTn search program (Altschul et al., 1990)

Results
We successfully implemented the ACP system to identify differentially expressed genes (DEGs) in larvae chronically exposed to cadmium, as previously demonstrated in blastocyst experiments (Cui et al., 2005;Hwang et al., 2004;Hwang et al., 2005). Our differential banding patterns of the cDNA representation of DEGs is summarized in Figure 1. Fourteen DEGs were identified after chronic exposure of An. gambiae larvae to cadmium heavy metal (Table 3). Most (11) of the differentially expressed genes were induced in cadmium exposed relative to the cadmium un-exposed controls. Our BLAST (REF) results revealed that the cadmium induced transcripts were clustered into metabolism (AGAP008584-RA, AGAP001249-RA and AGAP009563-RA), transport (AGAP012302-RA and AGAP002638-RA) and protein synthesis (AGAP028915-RA, AGAP004750-RA, AGAP028391-RA, AGAP003870-RA, AGAP028907-RA, AGAP028818-RA and AGAP028899-RA) processes.
Three of the DEGs identified were suppressed in the cadmium exposed larvae and these included AGAP006187-RA, AGAP002262-RA and AGAP003078-RA. The sequences were subsequently taken through a BLAST search. The results of the sequence analysis are shown on the manuscript.  The gel photo of a colony PCR of 20 samples that was carried out after blue/white colony screening using M13 primers.

Discussion
We identified ATP-binding cassette transporters belonging to the superfamily of membrane proteins that are present in all living organisms (Dean & Annilo, 2005). They are normally associated with movement of substrates such as amino acids, peptides, sugars, metals, inorganic ions, lipids, lipopolysaccharides and xenobiotics across biological membranes (Dawson & Locher, 2006;Hollenstein et al., 2007a). The ABC transporters have been shown to affect development, metabolism and insecticide resistance in insects (Borycz et al., 2008;Dow & Davies, 2006;Ricardo & Lehmann, 2009;Vache et al., 2007). The silencing of the ABCH1 gene has been shown to result in the death of larvae and pupae (Guo et al., 2015). Therefore, induction of the ABC transporters may suggest that they are involved in cadmium transport through membranes to reduce toxicity of cadmium metal to the larvae in their environment.
The induction of the eupolytin gene may have a role in the activation of defense molecules. In a study involving the ground beetle Eupolyphaga sinensis, eupolytin-1 gene encoding a protease was shown to be involved in the activation of plasminogen and hydrolysis of fibrinogen (Yang et al., 2011).
Ribosomal genes are involved in protein synthesis and upregulation of the various arrays of ribosomal RNAs in this study, which suggests their role in enhancing the survival of An. gambiae in the heavy metal polluted environment by the transcription and translation of genes which are important in the adaptation of the larvae to xenobiotics.
The nuclear structure referred to as THO complex is usually conserved in all kingdoms, and it has an important role in the packing of pre-mRNA molecules into RNA-protein assemblies which are termed mRNPs (Köhler & Hurt, 2007). A study carried out recently has shown that the THO complex is required for efficient expression of some genes, ensuring genetic stability thereby preventing transcription-associated recombination (Gewartowski et al., 2012). The expression of the THO complex is suggestive of its role in expressing defense genes that enhance survival of larvae in a Cadmium polluted environment.
Suppression of AGAP006187-RA, AGAP002262-RA and AGAP003078-RA transcripts that included G-Proteins couple receptors to adenylyl cyclase stimulation indicated increasing levels of cAMP and a cascade of events that constitute the signal transduction pathway that drive cellular responses. Metallopeptidases are a ubiquitous and diverse group of enzymes containing both endopeptidases and exopeptidases. Although they vary widely at the sequence, structural, and functional levels, they all require a metal ion for catalytic activity (Rawlings & Salvesen, 2013). The suppression of these important genes involved in signal transduction and proteolytic activity would account for the larval mortality rates that are usually observed in larvae raised in the cadmium heavy metal environment.
Our findings shed light on the adaptation of the An. gambiae mosquito to heavy metals by differentially expressing particular genes in response to a toxicant impact. A study to determine differentially expressed genes in cadmium-exposed sebastes schlegeli unraveled genes related to pathogenesis, extrinsic stresses, and catalytic metabolites (Woo & Yum, 2014). Previous studies have indicated that metallothionein and α-tubulin genes that are present in insects can be used as potential biomarkers (Hare, 1992;Klerks & Weis, 1987;Mattingly et al., 2001;Roesijadi, 1994). Metallothionein was assessed through C. quinquefasciatus mosquito larvae for Copper, Cadmium and Zinc aquatic environmental levels (Sarkar et al., 2004). Therefore, the genes identified might be useful in the development of potential biomarkers that can be used to assess the level of environmental pollution of heavy metal origin in An. gambiae mosquitoes.

Conclusions
We were able to identify genes that are differentially expressed due to chronic exposure of An. gambiae larvae to cadmium metal using the ACP-based PCR method. However, application of more sensitive techniques like those used in proteomics would generate more data.

Competing interests
No competing interests were disclosed.

Grant information
Funding for this study was provided by the Department of Research and Extension, Egerton University and the DAAD in-country Scholarship.
The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Acknowledgements
We hereby wish to acknowledge the following individuals for their contribution to this work: The Head of the Capacity Building Department at ICIPE, for granting us permission to carry out this work in their Molecular and Biotechnology unit.
The Director of the Research and Extension Department at Egerton University.
The DAAD team for the financial support, which enabled this work to be completed.

Open Peer Review
While the RNA extraction and initial cDNA synthesis were executed properly, I'm worried about the follow up steps and the verification of up-or down regulation.
Major Issues: Second stage PCR of 40 cycles may introduce tremendous bias in fragments that differ in PCR efficiency. After so many cycles only shorter fragments will be preferentially amplified. This may inpart explain why so few DEGs have been identified in this study. All other competing DEGs with lower efficiency will be lost.

○
The GeneFishing technique is outdated. I would rather recommend SSH, microarray analysis or RNAseq. All of these techniques would give a much more complete picture of gentic differentces in the Cd adapted populations.

○
It is essential to verify DEGs using for instance QPCR. I used to work with differential PCR screening, and quite a number of the initial leads were false positives. So, an independent second method should be used to validate the DEGs ○ Most of the identified DEGS are associated with rRNA indicating that the dT-ACP 1 primer was not specific enough to target mRNA. This is very worrying and confirms my 1st issue raised.

○
In conclusion, these are very preliminary data generated with an out dated system.
Is the work clearly and accurately presented and does it cite the current literature?