ORIGINAL PAPER

Comparative analysis of midgut bacterial communitiesof Aedes aegypti mosquito strains varyingin vector competence to dengue virus

Shakti S. Charan & Kiran D. Pawar &

David W. Severson & Milind S. Patole &

Yogesh S. Shouche

Received: 2 August 2012 /Accepted: 5 April 2013# Springer-Verlag Berlin Heidelberg 2013

Abstract Differences in midgut bacterial communities of Ae-des aegypti, the primary mosquito vector of dengue viruses(DENV), might influence the susceptibility of these mosqui-toes to infection by DENV. As a first step toward addressingthis hypothesis, comparative analysis of bacterial communitiesfrom midguts of mosquito strains with differential geneticsusceptibility to DENV was performed. 16S rRNA gene li-braries and real-time PCR approaches were used to character-ize midgut bacterial community composition and abundance inthree Aedes aegypti strains: MOYO, MOYO-R, and MOYO-S. Although Pseudomonas spp.-related clones were predomi-nant across all libraries, some interesting and potentially sig-nificant differences were found in midgut bacterialcommunities among the three strains. Pedobacter sp.- andJanthinobacterium sp.-related phylotypes were identified onlyin the MOYO-R strain libraries, while Bacillus sp. wasdetected only in the MOYO-S strain. Rahnella sp. was found

in MOYO-R and MOYO strains libraries but was absent inMOYO-S libraries. Both 16S rRNA gene library and real-timePCR approaches confirmed the presence of Pedobacter sp.only in the MOYO-R strain. Further, real-time PCR-basedquantification of 16S rRNA gene copies showed bacterialabundance in midguts of the MOYO-R strain mosquitoes tobe at least 10–100-folds higher than in the MOYO-S andMOYO strain mosquitoes. Our study identified some putativebacteria with characteristic physiological properties that couldaffect the infectivity of dengue virus. This analysis representsthe first report of comparisons of midgut bacterial communitieswith respect to refractoriness and susceptibility of Aedesaegypti mosquitoes to DENV and will guide future efforts toaddress the potential interactive role of midgut bacteria ofAedes aegypti mosquitoes in determining vectorial capacityfor DENV.

Introduction

Dengue viruses (DENV) cause significant morbidity and mor-tality worldwide and are transmitted to humans through thebites of infective female Aedes aegypti mosquitoes. Globalincidences of dengue have grown dramatically in recent de-cades, and at present, DENV is endemic in 112 countriescausing around 50–100 million infections and 24,000 deathsannually (WHO1997, 2009; Halstead 1999). Given the globalre-emergence of mosquito borne diseases, there is an urgentneed to explore alternative strategies like genetic modificationof mosquitoes and paratransgenesis approaches to combat thisdrastic situation (Ito et al. 2002; Riehle et al. 2007).

Since the first point of contact between a pathogen ingestedwith the blood meal and the mosquito is the midgut epithelialsurface, a pathogen must overcome midgut infection barriers

Electronic supplementary material The online version of this article(doi:10.1007/s00436-013-3428-x) contains supplementary material,which is available to authorized users.

S. S. Charan :K. D. Pawar :M. S. Patole :Y. S. Shouche (*)Molecular Biology Unit, National Centre for Cell Science,Ganeshkhind, Pune, Maharashtra, Indiae-mail: yogesh@nccs.res.in

S. S. Charane-mail: sscharan17@gmail.com

K. D. Paware-mail: pawarkiran1912@gmail.com

M. S. Patolee-mail: patole@nccs.res.in

D. W. SeversonEck Institute for Global Health Division of Biological Sciences,University of Notre Dame, Notre Dame, IN, USAe-mail: severson.1@nd.edu

Parasitol ResDOI 10.1007/s00436-013-3428-x

such as digestive enzymes, lectins, antimicrobial peptides,nitric oxide, and the prophenoloxidase complex (Ratcliffeand Whitten 2004; Michel and Kafatos 2005). Along withthese, naturally occurring midgut bacteria might also play avital role in infection outcome as, for example, midgut bacte-ria were reported to inhibit sporogonic development of Plas-modium in Anopheles (Pumpuni et al. 1993, 1996; Gonzalez-Ceron et al. 2003). Midgut bacteria were also reported toinfluence the parasitic life cycle in other insect vectors suchas sand flies and tsetse flies (Schlein et al. 1985; Welburn andMaudlin 1999). Further, there is considerable interest in char-acterization of endogenous microbiota in the mosquito midgutto identify potential bacteria for paratransgenesis approachesto infection prevention (Riehle and Jacobs-Lorena 2005;Riehle et al. 2007). Earlier, Favia et al. (2007) reported thatbacteria belonging to the genus Asaia can stably associatewith Anopheles stephensi mosquitoes and represent potentialcandidates for paratransgenesis approach for expressing anti-parasite effector molecules. Recent studies on mosquito mid-gut bacteria confirm a growing interest in the research com-munity toward this strategy (Rani et al. 2009; Gusmão et al.2010; Chavshin et al. 2012; Dinparast Djadid et al. 2011;Terenius et al. 2012; Wang et al. 2012).

Based on this evidence, we explored the potential that thebacteria present in the Aedes aegypti midgut may play impor-tant roles in determining successful DENV midgut infectionthat would, therefore, influence vector competence. Here, weinvestigated differences in midgut microbial communitiesamong three Aedes aegypti laboratory strains with contrastingsusceptibility to DENV infection. Although the contrastingabilities of these Aedes aegypti strains to transmit differentpathogens has been correlated with complex genetic factors(Morlais and Severson 2001; Morlais et al. 2003; Yan andSeverson 2003), the role of naturally occurring mosquitomidgut bacteria in the observed phenotypic variability insusceptibility phenotype is unknown. PCR amplicons of bac-terial 16S rRNA genes obtained from midgut preparations ofeach of the three Aedes aegypti strains were cloned, se-quenced, and compared qualitatively and quantitatively by16S rRNA gene library analysis and real-time PCR.

Materials and methods

Mosquito rearing

Three Aedes aegypti strains were compared in this study; theMOYO strain was originally collected from East Kenya in1974, while the MOYO-R andMOYO-S strains were initiallydifferentially selected from the MOYO strain for high refrac-toriness and high susceptibility, respectively, to Plasmodiumgallinaceum (Thathy et al. 1994) and also subsequently deter-mined to show similar contrasting susceptibilities to DENV

infection (Schneider et al. 2007). These three strains werereared and maintained in the same walk-in environmentalchamber at the University of Notre Dame since 2003. Thedetailed protocol for mosquito culture is described elsewhere(Clemons et al. 2010). Briefly, mosquitoes were reared at26 °C, 84 % relative humidity, under a 16:8-h light/dark cyclewith a 1-h crepuscular period at the beginning and at the endof each light cycle. Larvae were hatched in ~2 L tepid waterand given an ad libitum solution of bovine liver powder (ICNBiomedicals, Inc, Irvine, CA, USA). Adults were maintainedon a 5 % sucrose solution ad libitum. The water, bovine liverpowder solution, and sucrose solution were each provided tostrains from a common source. Female mosquitoes wereblood-fed on anesthetized rats.

DNA isolation

Adult female mosquitoes were surface-sterilized in 70 %ethanol for 5–10 min. The midguts of adult mosquitoes werethen carefully dissected under sterile conditions. Five mid-guts for each strain were pooled and suspended in 0.85 %NaCl, followed by homogenization with a plastic pestle in1.5 ml Eppendorf tubes and subjection to DNA extractionsas previously described (Severson 1997). DNA quality waschecked on 0.8 % agarose (USB, USA) gels for purity andquantified using NanoDrop (Thermo Scientific, USA).

PCR amplification of 16S rRNA genes and construction ofclone libraries

For this study, two sets of universal bacterial primers wereused for construction of 16S rRNA gene libraries to capturemaximum microbial diversity and to remove any potentialPCR bias caused by primer set (Suzuki and Giovannoni1996; Polz and Cavanaugh 1998). Total midgut DNA iso-lates were used as template in PCRs employing universalbacterial 16S rRNA gene specific primers: (1) primer set A:530F/1490R (Weisburg et al. 1991) and (2) primer set B:357F/1492R (Lane et al. 1985). PCR conditions were gen-erally as previously described (Wani et al. 2006), with thenumbers of cycles per reaction reduced to 25 to minimizePCR bias (Suzuki and Giovannoni 1996; Polz andCavanaugh 1998). The PCR products were purified usinga PCR purification kit (Qiagen, USA), ligated into theTOPO pCR®2.1 vector (Invitrogen, USA), and thentransformed into One Shot® TOP10 chemically competentEscherichia coli cells (Invitrogen, USA) following the man-ufacturer’s instructions. PCR screening of individual cloneswas performed and positive clones were sequenced usingM13 vector primers. Sequencing was performed on an ABIPRISM 3730 DNA Analyzer (Applied BioSystems, USA)using the ABI Big-Dye version 3.1 sequencing kit (AppliedBioSystems, USA) as per the manufacturer’s instructions.

Parasitol Res

Sequences from this study were deposited in GenBankunder accession numbers HQ873675–HQ873696.

Sequence assembly and phylogenetic analysis

Sequence assembly and editing was performed usingChromasPro 1.34 (www.technelysium.com.au/ChromasPro.html). For preliminary identifications, the 16S rRNA genesequences were analyzed using BLASTn (http://www.ncbi.nlm.nih.gov/BLAST/) and the Ribosomal Data-base Project II (RDP II) (http://rdp.cme.msu.edu) using theRDP query program (Maidak et al. 1999). The presence ofchimeric sequences was checked using the Pintail 1.01 mod-ule of the Mallard 1.02 program (Ashelford et al. 2006), andpredicted chimeras were further analyzed by Bellerophon(Huber et al. 2004). Putative chimeric sequences were exclud-ed from further analysis. Multiple sequence alignment of allsequences in each 16S rRNA gene library was performedusing ClustalX 1.83 (Thompson et al. 1994), and alignedsequences were edited manually using DAMBE (Xia andXie 2001) to obtain an unambiguous sequence alignment.Distance matrices were constructed using the DNAdist mod-ule of Phylip 3.64 (Felsenstein 1989).

Operational taxonomic unit determination and bacterialdiversity analysis

The resulting distance matrices were used as input in theDOTUR 1.53 program for calculating various diversity in-dices (Schloss and Handelsman 2005). The rarefaction anal-yses, phylotype richness, evenness (Pielous’ index),Shannon diversity index (H), and Simpson index were cal-culated for operational taxonomic units (OTUs) with anevolutionary distance (D) of 0.03 (or about 97 % 16S rRNAgene sequence similarity). The sequences representing eachOTU were compared with the current database of nucleotidesequences at GenBank and at the Ribosomal Database Pro-ject (RDP), and the closest matches to bacterial strains wereobtained. Phylogenetic trees were constructed by the Neigh-bor Joining method using Kimura 2 parameter distances inMEGA 4.0 software (Tamura et al. 2007).

Real-time PCR assay for bacterial quantification

16S rRNA gene copy numbers in midguts of the three mos-quito strains were determined by real-time PCR. Each reactionwas carried out in duplicate by using TaqMan Universal PCRmaster mix (Applied Biosystems) in ABI 7300 Real TimePCR system (Applied Biosystems). 16S rRNA gene-basedbacterial quantification was done by using 331F and 797RPrimers (numbering based on E. coli 16S rRNA gene) and(6-FAM)-5 ′ -CGTATTACCGCGGCTGCTGGCAC-3′-(TAMRA) probe, as described previously (Nadkarni et al.

2002). Purified Pseudomonas aeruginosa DNA was used in1:10 dilution series (10 pg–100 ng) for generating a standardcurve. Mass of the Pseudomonas aeruginosa genome havingsingle copy of 16S rRNA genewas calculated according to thefollowing formula:

M ¼ ðnÞ 1:096 e� 21 g bp=ð Þ� 16S rRNA gene copies per genome;

where n = genome size.The CT values (i.e., the cycle no. in which exponential

amplification of PCR products crosses threshold) were de-termined on the basis of the fluorescence signals at the meanbaseline during the early cycles of amplification. PCR effi-ciency was calculated on the basis of standard by usingfollowing equation:

Efficiency ¼ 10 �1 slope=ð Þ � 1:

Real-time PCR-based quantification of Pedobacter

The probe and primer set were designed using the PrimerExpress Software (Applied Biosystems) and were based onconserved regions of identity within the 16S rRNA gene ofPedobacter sp. W48 (DQ778037). This software gave somebest fit suggestions for the probe and primer set which werechecked against the NCBI and RDP database for their speci-ficity to Pedobacter genus. The final chosen set included theprobe, (6-FAM)—5′-CCGCGTGCAGGAAGACAGCCC-3′(TAMRA), the forward primer, 5 ′-AATGGAGGCAACTCTGAACCA-3′, and the reverse primer, 5′-TCCCGAATAAAAGCAGTTTACGA-3′. Real-time PCR reactionswere performed using the ABI 7300 Real-Time PCR system(Applied Biosystems), and each reaction was carried out induplicates. Primer and probe concentrations were optimizedfor the ratio of 500:100 nM. Each PCR reaction wasperformed in a volume of 25 μl using Taqman universalPCRmaster mix supplied byApplied Biosystem. The reactionconditions were as follows: 50 °C for 2 min, 95 °C for 10 minand 40 cycles of 95 °C for 15 s and 60 °C for 1 min. Data wereanalyzed in Sequence Detection Software version 1.6.3 (Ap-plied Biosystems). Purified Pedobacter sp. W48 DNA wasused for generating standard curve and PCR efficiency, andgenome mass was calculated as described above.

Results

16S rRNA gene library analysis

Comparative analysis of clone libraries representing the mid-gut bacterial diversity associated with the MOYO-R, MOYO-S, and MOYO strains of Aedes aegypti constructed with

Parasitol Res

universal bacterial 16S rRNA gene-specific primers 530F and1490R (primer set A) showed some noteworthy differencesbetween midgut microbiota of the three strains (Table 1). InMOYO-R strain library, 196 clones were sequenced resultingin 135 high-quality sequences. Out of these, eight were foundto be chimeras and therefore were excluded from furtheranalysis. DOTUR analysis of the remaining clone library se-quences identified four OTUs phylogenetically affiliated withBacteriodetes, Gammaproteobacteria, and Betaproteobacteria(Fig. 1).Pseudomonas sp. (AB440177)-related sequences (OTUR02A) dominated the library since these constituted 72 % oftotal clones examined. OTU R17A (representing 10 % of totalclones) was most closely related (99 %) to the 16S rRNA geneof Rahnella sp. (U88436) of Enterobacteriaceae family. OTUR92A (representing 17 % of total clones) showed the greatestsimilarity with uncultured bacterium of Oxalobacteraceae fam-ily (AY958896), and the nearest phylogenetic neighbor wasJanthinobacterium sp. (96 % gene identity). A singleton OTUR71A showed maximum identity of 99 % with Pedobacter sp.(AJ438170).

In the MOYO-S strain library, sequencing of 196 clonesyielded 150 high-quality sequences. Out of these 145 werechosen for final analysis after exclusion of five chimericsequences. Phylogenetic analysis of these sequences re-vealed four OTUs affiliated with Betaproteobacteria,Gammaproteobacteria, and Firmicutes (Fig. 2). Pseudo-monas sp. (EU708624)-related sequences (OTU S47A)dominated the library with 62 % occurrence. Thirtyseven percent of total sequences representing OTUsS69A and S46A were related to uncultured bacteria ofAlcaligenaceae and Oxalobacteraceae families, whileOTU S21A showed 99 % identity to Bacillus sp.

(FJ210679). Sequences related to Enterobacteriaceaefamily were absent from this library.

In the MOYO strain, only bacteria belonging to classGammaproteobacteria were present (Fig. 3). Again, Pseu-domonas sp. (EU364532)-related clones were predominantin the library with 77 % occurrence in 109 total sequences.OTU M15A representing 3 % of total sequences showed thegreatest identity to uncultured bacterium (HM142075) ofEnterobacteriaceae family while OTU M11A (representing20 % of total sequences) showed identity with Rahnella sp.(EU181369). To capture maximum bacterial diversity, allthree 16S rRNA gene libraries associated with the threestrains under study were repeated with another primer set(357F/1492R or primer set B). These libraries again showeda near congruent composition of bacterial communities asPseudomonas-related clones were predominant in all ofthem (Table 2). The only notable result obtained was thepresence of Janthinobacterium sp. (AY958995)-related se-quences in MOYO-R strain library.

To study the phylogenetic relationship among the se-quences affiliated to Pseudomonas from the three Aedesaegypti strains, phylogenetic tree, based on 99 % sequencesimilarity, was constructed using all such sequences from sixlibraries and analyzed. It was observed that sequences gener-ated from MOYO-R strain were grouped together while se-quences from MOYO-S and MOYO strain showed anoverlapping pattern (Supplementary Fig. 1). BLASTn analy-sis of the sequences identified the presence of various strainsofPseudomonas such asPseudomonas libanensis strain Cl-12(KC178586) (R28A), Pseudomonas psychrophila strain HA-4 (JQ968688) (R27B), Pseudomonas brenneri (AM933521)(R64B), Pseudomonas brassicacearum (CP002585) (R83B),

Table 1 Abundance of bacteria within 16S rRNA gene libraries constructed using primer set A in Aedes aegypti mosquito strains

Division Family MOYO-R MOYO-S MOYO

OTUname

Closest databaserelative

OTUname

Closest databaserelative

OTUname

Closest databaserelative

Gammaproteobacteria

Pseudomonadaceae R02A(72)a Pseudomonas sp.(AB440177 )b

S47A(62) Pseudomonas sp. (EU708624) M46A(83) Pseudomonas sp.(EU364532)

Enterobacteriaceae R17A(10) Rahnella sp. (U88436) – M11A(14) Rahnella sp.(EU181369)

M15A(3) Uncultured gammabacterium clone(HM142075)

Betaproteobacteria

Oxalobacteraceae R92A(17) Uncultured bacterium clone(AY958896)

S69A(12) Uncultured bacterium clone(EU244073)

–

Alcaligenaceae – S46A(25) Uncultured bacteriumclone(DQ984533)

–

Bacteroidetes Sphingobacteriaceae R71A(1) Pedobacter sp. (AJ438170) – –

Fermicutes Bacillaceae – S21A(1) Bacillus sp. (FJ210679) –

OTU operational taxonomic unita Values in parenthesis correspond to percentage distribution of the respective operational taxonomic unit in the 16S rRNA gene libraryb GenBank accession number (in parenthesis) of the closest database relative of respective OTU

Parasitol Res

and Pseudomonas trivialis strain P 513/19 (NR_028987)(R22A) only in MOYO R strain. This observation indicatedthat MOYO-R strain mosquitoes harbored the members ofPseudomonas which were different from that of MOYO-Sand MOYO strain mosquitoes.

The diversity index analysis indicated that most of thebacterial diversity was almost covered (Table 3). Consider-ing all the 16S rRNA gene libraries, Shannon–Weaver di-versity index varied from 1.01 to 1.28 and was slightlyhigher in MOYO-R strain than MOYO-S and MOYOstrains. Simpson diversity index varied from 0.37 to 0.60,and evenness index or Pielous’ index was between 0.69 and0.82. The percentage coverage of the sequence analysis wasalso calculated by estimating the Good’s coverage valuesusing Good’s method (Good 1953). The Good’s coveragevalues for MOYO-R strain library A and MOYO-S strainlibrary Awere estimated 99 %, while all remaining librariesshowed 100 % coverage value. Rarefaction curves from allthe six libraries were saturated indicating that our sequenc-ing efforts were exhaustive and all bacterial diversity wascaptured (Fig. 4).

Real-time PCR-based bacterial quantification

Mass of the Pseudomonas aeruginosa genome havingsingle copy of 16S rRNA gene for generating DNAstandard was calculated as 1.7 fg. DNA standard for16S rRNA gene quantification showed a slope of −3.15which corresponded to about 107 % PCR efficiency(Fig. 5). Real-time PCR-based quantification of totalbacteria showed noteworthy differences in copy no. of16S rRNA gene from three mosquito strains under stud-y. The copy numbers of 16S rRNA gene ranged from106 to 108 copies and were the highest in midgut ofMOYO-R strain (6.88×108) followed by MOYO andMOYO-S strains (4.36×107 and 1.44×106).

Determination of Pedobacter 16S rRNA gene copy number

Real-time PCR-based quantification of Pedobacter sp. indicatedthat these bacteria were present only in the midgut of theMOYO-R strain. We observed no amplification from otherbacterial cultures included in the PCR reaction indicating the

Fig. 1 Neighbor-joining tree constructed from 16S rRNA gene librarysequences from MOYO-R strain Aedes aegypti mosquitoes. The treewas generated by using the neighbor-joining method with Kimura 2parameter distances in MEGA 4.0 software. Numbers at nodes indicate

percent bootstrap values above 50 (1,000 replicates). The bar indicatesthe Jukes–Cantor evolutionary distance. Names in bold representOTUs obtained in this study. Accession numbers of the nearest neigh-bors are given in parentheses

Parasitol Res

specificity of probe toPedobactergenus.Mass of thePedobactersp. W48 genome having single copy of 16S rRNA gene forgenerating a DNA standard curve was calculated as 1.9 fg, andthe DNA standard curve showed a slope of −3.35 correspondingto about 99 % PCR efficiency (Fig. 5). Pedobacter sp. 16SrRNA gene copy numbers were 1.1×107 copies in MOYO-Rstrain contributing to about 2 % of total bacteria.

Discussion

In this study, we explored the midgut bacteria of lab strainsof Aedes aegypti with differential susceptibility to DENV.Our results are particularly interesting in that we have pre-viously shown that midgut bacteria of Aedes aegypti mos-quitoes have a significant effect on DENV infection successand likely play important roles in determining their vectorialcapacity to transmit the virus (Mourya et al. 2002a, b).Earlier Xi et al. (2008) also reported that naturally occurringmidgut bacteria of Aedes aegypti mosquitoes can modulatedengue virus infection by activating the Toll immune path-way, but could not find a direct bacteria–virus interactionand raised the possibility that mosquito midgut bacteria can

influence virus interaction with the midgut epithelium.However, as every bacterium has its unique niche in midgutenvironment, this interaction may vary from bacterium tobacterium. To determine whether population richness orpresence/absence of key bacteria is correlated to refractori-ness or susceptibility of Aedes aegyptimosquitoes to DENV,we used real-time PCR and 16S rRNA gene library ap-proaches. Since it has been estimated that almost 99 % ofall microorganisms cannot be cultivated by standardmethods (Hugenholtz et al. 1998) and available convention-al culture techniques limit the isolation and identification ofall the microorganisms of mosquito midgut, we choseculture-independent analysis to assess microbial diversityin Aedes aegypti midgut. The dominant bacteria in the mid-guts of these three mosquito strains belonged to Pseudomo-nas. Although not as dominant member, few previousstudies have also reported the presence of Pseudomonas inthe midguts of other mosquitoes such as Culex and Anoph-eles (Gonzalez-Ceron et al. 2003; Pidiyar et al. 2004; Riehleand Jacobs-Lorena 2005; Favia et al. 2007). Our finding ofPseudomonas being the dominant bacterium in the midgut isin congruence with the previous finding by Chavshin et al.(2012) that also reported Pseudomonas species as the most

Fig. 2 Neighbor-joining tree constructed from 16S rRNA gene librarysequences from MOYO-S strain Aedes aegypti mosquitoes. The treewas generated by using the neighbor-joining method with Kimura 2parameter distances in MEGA 4.0 software. Numbers at nodes indicate

percent bootstrap values above 50 (1,000 replicates). The bar indicatesthe Jukes–Cantor evolutionary distance. Names in bold representOTUs obtained in this study. Accession numbers of the nearest neigh-bors are given in parentheses

Parasitol Res

dominant members in midguts of Anopheles stephensi mos-quito at larval and adult stage. As we further analyzed thePseudomonas-related sequences to see the strain level var-iations, we observed that MOYO-R strain mosquitoes wereharboring different strains of Pseudomonas than of MOYO-S and MOYO strain mosquitoes. The exclusive presence ofthese Pseudomonas strains may be having a role to play indetermining the vector competence of the MOYO-R strain.

Furthermore, it was also shown that incorporation ofPseudomonas isolates in the mosquito blood meal resultsin an increased susceptibility of Culex quinquefasciatusmosquitoes to Japanese encephalitis virus (Mourya et al.2002a). Rahnella, an Enterobacteriaceae family member,was identified in MOYO-R and MOYO strains. It has alsobeen reported from the midgut of arbovirus vectorCulicoides sonorensis (Campbell et al. 2004).

In MOYO-R strain, the most interesting finding was thepresence of Pedobacter and Janthinobacterium. Pedobacteris commonly found in water and soil and also found to beassociated with scorpion intestine and a plant parasite nem-atode (Wang et al. 2007; Tian et al. 2011). It is a heparinase(EC 4.2.2.7) secreting bacterium which degrades heparansulfate, an important receptor moiety used by many patho-gens like Plasmodium, viruses, Trypanosoma cruzi, andLeishmania for binding to cell surface (Love et al. 1993;Herrera et al. 1994; Pinzon-Ortiz et al. 2001; Germi et al.2002). Recently, Sinnis et al. (2007) observed similaritiesbetween mosquito heparan sulfate and human liver heparan

sulfate. In the study, the authors postulated the importanceof this moiety for the transmission of Plasmodium in mos-quito host and also raised the possibility that other mosquitoborne pathogens like DENV may utilize mosquito heparansulfate for their survival in host mosquitoes. It was alsoobserved that heparin lyase can significantly inhibit thebinding of dengue virus and yellow fever virus by degradingthe heparin sulfate (Germi et al. 2002). Having the potentialof degrading heparan sulfate, Pedobacter might be playingan important role in determining the refractoriness of thesemosquitoes to DENV and Plasmodium and also could havepotential in developing a novel strategy to control insecttransmitted diseases.

Janthinobacterium is commonly found in rivers, lakes,and springs and has also been reported from midgut ofeonymph of saw fly (Zahner et al. 2008). Janthinobacteriumproduces a water-insoluble pigment violescin, which wasshown to possess anti-bacterial, anti-viral, anti-leishmanial,and anti-tumoral activities (Dessaux et al. 2004). The mem-bers of Janthinobacterium may have role to play in devel-oping immunity of mosquitoes to pathogens. In the view ofthese unique physiological properties of Pedobacter andJanthinobacterium, here we assume that the presence ofthese bacteria in midgut of MOYO-R strain mosquitoescould be associated with refractoriness of these mosquitoesto DENV and Plasmodium.

In the MOYO-S strain, the noteworthy difference was thepresence of Bacillus sp. and absence of Enterobacteriaceae.

Fig. 3 Neighbor-joining tree constructed from 16S rRNA gene librarysequences from MOYO strain Aedes aegypti mosquitoes. The tree wasgenerated by using the neighbor-joining method with Kimura 2 pa-rameter distances in MEGA 4.0 software. Numbers at nodes indicate

percent bootstrap values above 50 (1,000 replicates). The bar indicatesthe Jukes–Cantor evolutionary distance. Names in bold representOTUs obtained in this study. Accession numbers of the nearest neigh-bors are given in parentheses

Parasitol Res

Bacillus sp. has been previously reported from midguts ofCulex quinquefasciatus, Anopheles stephensi, and Aedesaegypti mosquitoes (Pumpuni et al. 1996; Gonzalez-Ceronet al. 2003; Pidiyar et al. 2004). An intermediate display ofbacterial diversity of above two strains was seen in MOYOstrain with presence of Enterobacteriaceae and absence ofPedobacter, Janthinobacterium, and Bacillus sp.

In our study, diversity indices analysis showed lowvalues of Shannon–Weaver diversity index and evennessindex in all the six clone libraries predicting very lowbacterial diversity in Aedes aegypti midgut environment.There was not much difference in diversity indices betweenthe clone libraries of the three mosquito strains indicating

same low levels of bacterial diversity. This could possiblyindicate to a very stringent mosquito midgut environmentthat supports minimal bacterial colonization.

In RT-PCR analysis, we used Pseudomonas aeruginosagenomic DNA for generating standard as it was postulatedearlier that real-time PCR standards should be constructedwith the bacteria that predominate a given habitat so as tominimize variations caused by differences in 16S rRNAgene copy numbers and bacterial generation time (Nadkarniet al. 2002). Real-time PCR-based bacterial quantificationshowed that 16S rRNA gene copy numbers in MOYO-Rstrain were higher than MOYO and MOYO-S strains. It wasobserved that reduction in normal gut flora of Anopheles

Table 2 Abundance of bacteria within 16S rRNA gene libraries constructed using primer set B in Aedes aegypti mosquito strains

Division Family MOYO-R MOYO-S MOYO

OTUname

Closest databaserelative

OTUname

Closest databaserelative

OTUname

Closest databaserelative

Gammaproteobacteria

Pseudomonadaceae R48B(58)a Pseudomonas sp.(AF074384)b

S56B(20) Pseudomonas sp.(FJ179369)

M40B(77) Pseudomonas sp. (EU864269)

R101B(8) Uncultured bacteriumclone (AY959027)

S37B(60) Pseudomonas sp.(EF100617)

M03B(5) Uncultured bacterium clone(AY959000 )

Enterobacteriaceae R03B(24) Rahnella sp. (U90757) – M97B(15) Rahnella sp. (DQ885948)

M78B(3) Uncultured bacterium clone(EU024329)

Betaproteobacteria

Oxalobacteraceae R90B(10) Janthinobacterium sp.(AY958995)

– –

Alcaligenaceae – S48B(20) Unculturedbacteriumclone(DQ354715)

–

OTU operational taxonomic unita Values in parenthesis correspond to percentage distribution of the respective operational taxonomic unit in the 16S rRNA gene libraryb GenBank accession number (in parenthesis) of the closest database relative of respective OTU

Table 3 Comparison of the phylotype richness, Shannon index, evenness values (Pielous’ index), and Simpson index of 16S rRNA gene librariesfrom MOYO-R, MOYO-S, and MOYO strain Aedes aegypti mosquitoes

Index MOYO-R MOYO-S MOYO

Primer set A Primer set B Primer set A Primer set B Primer set A Primer set B

No. of clonesa 127 95 145 99 109 105

Phylotype richnessb 4 4 4 3 3 4

Shannon index (H)c 1.24 1.28 1.15 1.03 1.01 1.17

Evenness (E)d 0.78 0.82 0.76 0.72 0.69 0.75

Simpson indexe 0.42 0.37 0.45 0.54 0.60 0.52

Numbers were calculated with DOTUR program and OTUs were defined using a distance level of 3 %a Total numbers of sequences taken for final analysisb Total numbers of OTUs in the 16S rRNA gene libraryc The Shannon–Weaver diversity index (H). H was calculated as follows: H ¼ P

pið Þ log2 p� 1ð Þ, where p represents the proportion of a distinctphylotype relative to the sum of alld Evenness/Pielous’ index (E) was calculated as: E ¼ H log2= ðSÞ where log2ðSÞ ¼ Hmaxe Simpson coverage index

Parasitol Res

mosquitoes increases their susceptibility toward Plasmodi-um infection (Beier et al. 1994) and our results are inagreement with the same. To the best of our knowledge,this is the first attempt to quantify the mosquito midgutbacteria using real-time Taqman PCR-based technique. Toconfirm the clone library results, we also rechecked thepresence of Pedobacter bacteria with real-time TaqmanPCR approach and confirmed its presence only in MOYO-R strain. When the abundance of these bacteria in mosquitomidgut was estimated, it was found that about 2 % of thetotal bacteria were constituted by Pedobacter. This is alsothe first report of quantification of these bacteria with real-time PCR approach with highly specific Taqman probe, and

as these bacteria are used for commercial production ofheparinase enzyme (Lohse and Linhardt 1992), this reportmay also prove valuable in this aspect.

Based on RT-PCR and 16S rRNA gene library analysis,we conclude that there were some interesting differencesbetween midgut bacterial populations of MOYO-R,MOYO-S, and MOYO mosquitoes which could play impor-tant roles in refractoriness and sensitivity of these mosqui-toes to dengue virus. Our comparative analysis of midgutmicrobial communities of these strains would serve as atemplate for future research on the potential role of midgutmicrobial communities in mosquito competence. Any vari-ations in midgut bacteria identified in this study will need tobe analyzed further in subsequent experiments to robustlycorrelate gut microbiota differences and variations in den-gue virus infectivity. Whether the differences in midgutbacterial communities reported here is maintained betweenAedes aegypti strains in other habitats and can be directlycorrelated with vector competence also need to be seriouslyaddressed in future studies. Importantly, our findings mightcontribute to the identification of new targets for anti denguestrategies and could influence design of novel approaches toeradicate mosquito-borne diseases.

References

Ashelford KE, Chuzhanova NA, Fry JC, Jones AJ, Weightman AJ(2006) New screening software shows that most recent large 16SrRNA gene clone libraries contain chimeras. Appl EnvironMicrobiol 72:5734–5741

Beier MS, Pumpuni CB, Bizio JC, Davis JR (1994) Effect ofparaaminobenzenoic acid, insulin and gentamicin on Plasmodiumfalciparum development in anopheline mosquitoes (Diptera:Culicidae). J Med Entomol 31:561–565

Campbell CL, Mummey DL, Schmidtmann ET, Wilson WC (2004)Culture-independent analysis of midgut microbiota in the arbovirusvector Culicoides sonorensis (Diptera: Ceratopogonidae). J MedEntomol 41:340–348

Chavshin AR, Oshaghi MA, Vatandoost H, Pourmand MR, Raeisi A,Enayati AA, Mardani N, Ghoorchian S (2012) Identification ofbacterial microflora in the midgut of the larvae and adult of wildcaught Anopheles stephensi: a step toward finding suitableparatransgenesis candidates. Acta Trop 121:129–134

Clemons A, Mori A, Haugen M, Severson D, Duman-Scheel M (2010)Aedes aegypti culturing and egg collection. Cold Spring HarbProtoc 2010:pdb.prot5507

Dessaux Y, Elmerich C, Faure D (2004) Violacein: a molecule ofbiological interest originating from the soilborne bacteriumChromobacterium violaceum. Rev Med Interne 25:659–662

Dinparast Djadid N, Jazayeri H, Raz A, Favia G, Ricci I, Zakeri S(2011) Identification of the midgut microbiota of An. stephensiand An. maculipennis for their application as a paratransgenic toolagainst malaria. PLoS One 6:e28484

Favia G, Ricci I, Damiani C, Raddadi N, Crotti E, Marzorati M, RizziA, Urso R, Brusetti L, Borin S, Mora D, Scuppa P, Pasqualini L,Clementi E, Genchi M, Corona S, Negri I, Grandi G, Alma A,Kramer L, Esposito F, Bandi C, Sacchi L, Daffonchio D (2007)Bacteria of the genus Asaia stably associate with Anopheles

Fig. 4 Rarefaction curves generated for six 16S rRNA gene librariesof MOYO-R, MOYO-S, and MOYO strain Aedes aegypti mosquitoes

Fig. 5 Standard curves generated for the real-time PCR-based quanti-fication of total bacteria and Pedobacter sp. Square data points stan-dard curve for quantification of total bacteria; triangle data pointsstandard curve for quantification total bacteria. Standard curves wereobtained by plotting threshold cycle values (CT) as a linear function ofthe log 10 of the apparent amount of Pseudomonas aeruginosa DNA(for total bacteria) and Pedobacter sp. W48 DNA (for Pedobacter sp.)

Parasitol Res

stephensi, an Asian malarial mosquito vector. Proc Natl Acad SciUSA 104:9047–9051

Felsenstein J (1989) PHYLIP—phylogeny inference package (version3.2). Cladistics 5:164–166

Germi R, Crance JM, Garin D, Guimet J, Lortat-Jacob H, Ruigrok RW,Zarski JP, Drouet E (2002) Heparan sulfate-mediated binding ofinfectious dengue virus type 2 and yellow fever virus. Virology292:162–168

Gonzalez-Ceron L, Santillan F, Rodriguez MH, Mendez D,Hernandez-Avila JE (2003) Bacteria in midguts of field-collected Anopheles albimanus block Plasmodium vivax sporo-gonic development. J Med Entomol 40:371–374

Good IJ (1953) The population frequencies of species and the estima-tion of population parameters. Biom 40:237–264

Gusmão DS, Santos AV, Marini DC, Bacci M Jr, Berbert-Molina MA,Lemos FJ (2010) Culture-dependent and culture-independentcharacterization of microorganisms associated with Aedes aegypti(Diptera: Culicidae) (L.) and dynamics of bacterial colonization inthe midgut. Acta Trop 115:275–281

Halstead SB (1999) Is there an inapparent dengue explosion? Lancet353:1100–1101

Herrera EM, Ming M, Ortega-Barria E, Pereira ME (1994) Mediationof Trypanosoma cruzi invasion by heparan sulfate receptors onhost cells and penetrin counter-receptors on the trypanosomes.Mol Biochem Parasitol 65:73–83

Huber T, Faulkner G, Hugenholtz P (2004) Bellerophon: a program todetect chimeric sequences in multiple sequence alignments.Bioinformatics 20:2317–2319

Hugenholtz P, Goebel BM, Pace NR (1998) Impact of culture-independent studies on the emerging phylogenetic view of bacterialdiversity. J Bacteriol 180:4765–4774

Ito J, Ghosh A, Moreira AL, Wimmer EA, Jacobs-Lorena M (2002)Transgenic anopheline mosquitoes impaired in transmission of amalaria parasite. Nature 417:452–455

Lane DJ, Pace B, Olsen GJ, Stahl DA, Sogin ML, Pace NR (1985)Rapid determination of 16S ribosomal RNA sequences for phy-logenetic analyses. Proc Natl Acad Sci USA 82:6955–6959

Lohse DL, Linhardt RJ (1992) Purification and characterization ofheparin lyases from Flavobacterium heparinurn. J Biol Chem267:24347–24355

Love DC, Esko JD, Moser DM (1993) A heparin-binding activityon Leishmania amastigotes which mediates adhesion to cellularproteoglycans. J Cell Biol 123:759–766

Maidak BL, Cole JR, Parker CT Jr, Garrity GM, Larsen N, Li B,Lilburn TG, McCaughey MJ, Olsen GJ, Overbeek R, PramanikS, Schmidt TM, Tiedje JM, Woese CR (1999) A new version ofthe RDP (Ribosomal Database Project). Nucleic Acids Res27:171–173

Michel K, Kafatos FC (2005) Mosquito immunity against Plasmodium.Insect Biochem Mol Biol 35:677–689

Morlais I, Severson DW (2001) Identification of a polymorphic mucin-like gene expressed in the midgut of the mosquito, Aedes aegypti,using an integrated bulked segregant and differential displayanalysis. Genetics 158:1125–1135

Morlais I, Mori A, Schneider JR, Severson DW (2003) Targetedapproach toward identification of candidate genes determiningPlasmodium gallinceum susceptibility in Aedes aegypti. MolGen Genomics 269:753–764

Mourya DT, Gokhale MD, Pidiyar VJ, Barde PV, Patole MS, MishraAC, Shouche YS (2002a) Study of the effect of the midgutbacterial flora of Culex quinquefasciatus on the susceptibility ofmosquitoes to Japanese encephalitis virus. Acta Virol 46:257–260

Mourya DT, Pidiyar VJ, Patole MS, Gokhale MD, Shouche YS(2002b) Effect of midgut bacterial flora of Aedes aegypti on thesusceptibility of mosquitoes to dengue viruses. Dengue Bull26:190–194

Nadkarni MA, Martin FE, Jacques NA, Hunter N (2002) Determina-tion of bacterial load by real-time PCR using a broad-range(universal) probe and primers set. Microbiology 148:257–266

Pidiyar VJ, Jangid K, PatoleMS, Shouche YS (2004) Studies on culturedand uncultured microbiota of wild Culex quinquefasciatusmosquitomidgut based on 16S ribosomal RNA gene analysis. Am J TropMedHyg 70:597–603

Pinzon-Ortiz C, Friedman J, Esko J, Sinnis P (2001) The binding of thecircumsporozoite protein to cell surface heparan sulfate proteo-glycans is required for Plasmodium sporozoite attachment totarget cells. J Biol Chem 27:26784–26791

Polz MF, Cavanaugh CM (1998) Bias in template-to-product ratios inmultitemplate PCR. Appl Environ Microbiol 64:3724–3730

Pumpuni CB, Beier MS, Nataro JP, Guers LD, Davis JR (1993)Plasmodium falciparum: inhibition of sporogonic developmentin Anopheles stephensi by Gram-negative bacteria. Exp Parasitol77:195–199

Pumpuni CB, Demaio J, Kent M, Davis JR, Beier JC (1996) Bacterialpopulation dynamics in three anopheline species: the impact onPlasmodium sporogonic development. Am J Trop Med Hyg54:214–218

Rani A, Sharma A, Rajagopal R, Adak T, Bhatnagar RK (2009)Bacterial diversity analysis of larvae and adult midgut microflorausing culture-dependent and culture-independent methods inlab-reared and field-collected Anopheles stephensi—an Asianmalarial vector. BMC Microbiol 9:96–118

Ratcliffe NA, Whitten MMA (2004) Vector immunity in microbe–vector interactions. In: Gillespie SH, Smith GL, Osbourn A(eds) SGM symposium 63: microbe–vector interactions invector-borne diseases. Cambridge University press, Cambridge,pp 199–262

Riehle MA, Jacobs-Lorena M (2005) Using bacteria to express anddisplay anti-parasite molecules in mosquitoes: current and futurestrategies. Insect Biochem Mol Biol 35:699–707

Riehle MA, Moreira CK, Lampe D, Lauzon C, Jacobs-Lorena M(2007) Using bacteria to express and display anti-Plasmodiummolecules in the mosquito midgut. Int J Parasitol 37:595–603

Schlein Y, Polacheck I, Yuval B (1985) Mycoses, bacterial infections andantibacterial activity in sandflies (Psychodidae) and their possiblerole in the transmission of leishmaniasis. Parasitology 90:57–66

Schloss PD, Handelsman J (2005) Introducing DOTUR, a computerprogram for defining operational taxonomic units and estimatingspecies richness. Appl Environ Microbiol 71:1501–1506

Schneider JR, Mori A, Romero-Severson J, Chadee DD, Severson DW(2007) Investigations of dengue-2 susceptibility and body sizeamong Aedes aegypti populations. Med Vet Entomol 21:370–376

Severson DW (1997) RFLP analysis of insect genomes. In: CramptonJM, Beard CB, Louis C (eds) The molecular biology ofinsect disease vectors. Chapman & Hall, London, pp 309–320

Sinnis P, Coppi A, Toida T, Toyoda H, Kinoshita-Toyoda A, Xie J,Kemp MM, Linhardt RJ (2007) Mosquito heparan sulfate and itspotential role in malaria infection and transmission. J Biol Chem282:25376–25384

Suzuki M, Giovannoni SJ (1996) Bias caused by template annealing inthe amplification mixtures of 16S rRNA genes by PCR. ApplEnviron Microbiol 62:625–630

Tamura K, Dudley J, Nei M, Kumar S (2007) MEGA4: MolecularEvolutionary Genetics Analysis (MEGA) software version 4.0.Mol Biol Evol 24:1596–1599

Terenius O, Lindh JM, Eriksson-Gonzales K, Bussière L, Laugen AT,Bergquist H, Titanji K, Faye I (2012) Midgut bacterial dynamicsin Aedes aegypti. FEMS Microbiol Ecol 80:556–565

Thathy V, Severson DW, Christensen BM (1994) Reinterpretation ofgenetics of susceptibility of Aedes aegypti to Plasmodiumgallinaceum. J Parasitol 80:705–712

Parasitol Res

Thompson JD, Higgins DG, Gibson TJ (1994) CLUSTALW: improv-ing the sensitivity of progressive multiple sequence alignmentthrough sequence weighting, position specific gap penalties andweight matrix choice. Nucleic Acids Res 22:4673–4680

Tian X, Cheng X, Mao Z, Chen G, Yang J, Xie B (2011) Compositionof bacterial communities associated with a plant–parasitic nema-tode Bursaphelenchus mucronatus. Curr Microbiol 62:117–125

Wang BJ, Liu Y, Jiang JT, Liu B, Liu SJ (2007) Microbial diversity inscorpion intestine (Buthus martensii Karsch). Wei Sheng Wu XueBao 47:888–893

Wang S, Ghosh AK, Bongio N, Stebbings KA, Lampe DJ, Jacobs-LorenaM (2012) Fighting malaria with engineered symbiotic bacteria fromvector mosquitoes. Proc Natl Acad Sci USA 109:12734–12739

Wani AA, Suraksai VP, Siddharth J, Raghavan RG, Patole MS, RanadeD, Shouche YS (2006) Molecular analysis of microbial diversityassociated with the Loanar soda lake in India: an impact crater inbasalt area. Res Microbiol 157:928–937

Weisburg WG, Barns SM, Pelletier DA, Lane DJ (1991) 16S ribosomalDNA amplification for phylogenetic study. J Bacteriol 173:697–703

Welburn SC, Maudlin I (1999) Tsetse–trypanosome interactions: ritesof passage. Parasitol Today 15:399–403

WHO—World Health Organization (2009) Dengue and denguehaemorrhagic fever. Fact sheet no. 117. http://www.who.int/mediacentre/factsheets/fs117/en/index.html

World Health Organisation (1997) Dengue hemorrhagic fever:diagnosis, treatment, prevention and control, 2nd edn.WHO, Geneva

Xi Z, Ramirez JL, Dimopoulos G (2008) The Aedes aegypti toll pathwaycontrols dengue virus infection. PLoS Pathog 4(7):e1000098

Xia X, Xie Z (2001) DAMBE: software package for data analysis inmolecular biology and evolution. J Hered 92:371–373

Yan G, Severson DW (2003) Dynamics of molecular markers linked tothe resistance loci in a mosquito-Plasmodium system. Genetics164:511–519

Zahner V, Lucarotti CJ, McIntosh D (2008) Application of 16S rDNA-DGGE and plate culture to characterization of bacterial communitiesassociated with the sawfly, Acantholyda erythrocephala (Hymenoptera,Pamphiliidae). Curr Microbiol 57:564–569

Parasitol Res