Drosophila anti-nematode and antibacterial immune regulators revealed by RNA-Seq

21 Pages·2015·2.7 MB·English

by  Julio C. Castillo

Castilloetal.BMCGenomics (2015) 16:519
DOI10.1186/s12864-015-1690-2
RESEARCH ARTICLE Open Access
Drosophila
anti-nematode and antibacterial
immune regulators revealed by RNA-Seq
Julio C. Castillo1,2, Todd Creasy3, Priti Kumari3, Amol Shetty3, Upasana Shokal1, Luke J. Tallon3
and Ioannis Eleftherianos1*
Abstract
Background: Drosophila melanogaster activates a variety of immuneresponses against microbial infections.
However, information onthe Drosophila immune response to entomopathogenic nematode infections is currently
limited. The nematode Heterorhabditis bacteriophorais an insect parasite that forms a mutualisticrelationshipwith
thegram-negative bacteriaPhotorhabdus luminescens. Following infection, the nematodes release thebacteria that
quickly multiply within the insect and produce severaltoxins that eventuallykill the host. Althoughwe currently
know that the insect immune system interacts with Photorhabdus, informationon interaction with thenematode
vector is scarce.
Results: Here wehave used next generation RNA-sequencingto analyze thetranscriptional profile of wild-type
adult flies infected byaxenicHeterorhabditis nematodes (lacking Photorhabdus bacteria), symbiotic Heterorhabditis
nematodes (carryingPhotorhabdusbacteria), and Photorhabdus bacteria alone. Wehaveobtainedapproximately
54millionreadsfromthedifferentinfectiontreatments.BioinformaticanalysisshowsthatinfectionwithPhotorhabdus
altersthetranscriptionofalargenumberofDrosophilagenesinvolvedintranslationalrepressionaswellinresponseto
stress.However,Heterorhabditisinfectionaltersthetranscriptionofseveralgenesthatparticipateinlipidhomeostasis
andmetabolism,stressresponses,DNA/proteinsythesisandneuronalfunctions.Wehavealsoidentifiedgenesinthe
flywithpotentialrolesinnematoderecognition,anti-nematodeactivityandnociception.
Conclusions:ThesefindingsprovidefundamentalinformationonthemoleculareventsthattakeplaceinDrosophila
uponinfectionwiththetwopathogens,eitherseparatelyortogether.Suchlarge-scaletranscriptomicanalysessetthe
stageforfuturefunctionalstudiesaimedatidentifyingtheexactroleofkeyfactorsintheDrosophilaimmuneresponse
againstnematode-bacteriacomplexes.
Keywords:Drosophila,Photorhabdus,Heterorhabditis,RNA-sequencing,Transcriptomics,Immunity,Infection,Parasitism
Background moleculesthatparticipateinthehostimmuneresponseto
Host innate immune responses are broadly conserved nematode infections remains incomplete [3]. Novel anti-
across many phyla [1]. The study of the interaction be- nematode immune responses in the host are likely to be
tweeninvertebratemodelhostsandpathogenicorganisms identified in model systems in which the host has a se-
provides insights into the mechanisms underlying patho- quencedgenomeandcanbegeneticallymanipulated.The
gen virulence and host immunity, and complements the common fruit fly, Drosophila melanogaster, with a vast
use of mammalianmodelsbyenabling whole-animal high numberofgeneticsandgenomicstoolsavailable,iswidely
throughput infection assays and genome wide transcrip- recognized as an outstanding model to analyze immune
tome analyses [2]. Despite impressive advances in the signaling pathways and elucidate the molecular and gen-
broad field of innate immunity, our understanding of the eticbasisofimmunedefensemechanisms[4–6].
The insect pathogenic nematode Heterorhabditis
*Correspondence:[email protected] bacteriophora is emerging as a promising parasitic
1InsectInfectionandImmunityLab,DepartmentofBiologicalSciences, organism for studying nematode pathogenicity and
InstituteforBiomedicalSciences,TheGeorgeWashingtonUniversity,
characterizing the function of novel host factors that
WashingtonDC20052,USA
Fulllistofauthorinformationisavailableattheendofthearticle
©2015Castilloetal.ThisisanOpenAccessarticledistributedunderthetermsoftheCreativeCommonsAttributionLicense
(http://creativecommons.org/licenses/by/4.0),whichpermitsunrestricteduse,distribution,andreproductioninanymedium,
providedtheoriginalworkisproperlycredited.TheCreativeCommonsPublicDomainDedicationwaiver(http://
creativecommons.org/publicdomain/zero/1.0/)appliestothedatamadeavailableinthisarticle,unlessotherwisestated.
Castilloetal.BMCGenomics (2015) 16:519 Page2of21
contribute to anti-nematode immune reactions [7, 8]. Our goal was to identify the number and nature of
Heterorhabditis nematodes form a mutually beneficial Drosophila genes that are differentially regulated upon
symbiotic relationship with the Gram-negative bac- infection with the nematodes and their associated bac-
teria of the Enterobacteriaceae family, Photorhabdus teria. We find that distinct types of genes are regulated
luminescens, which are found in the gut of the worms during infection of the fly by each of the two patho-
[9]. Heterorhabditis infective juvenile (IJ) worms belong gens. Therefore these results indicate that different sets
to an obligate stage in the nematode life cycle that is of genes are involved in the interaction between the fly
required for infection of the insect. This stage is and the nematodes or their bacteria, and that the fly
analogous to the Caenorhabditis elegans dauer stage employs distinct strategies to fight infection against Het-
and the developmentally arrested infective third stage erorhabditis nematode parasites and their mutualistic
larva (L3) of many mammalian parasitic nematodes Photorhabdus bacteria.
[10]. IJs gain entry to the insect through natural open-
ings or by penetrating the cuticle. Once inside, the IJ
resume development and expel Photorhabdus into the Results
hemolymph where the bacteria begin to divide. After HeterorhabditisnematodesandPhotorhabdusbacteria
2-3 days of bacterial growth the insect succumbs to producedistincttranscriptomicprofilesinDrosophila
the infection with the concomitant conversion of the We generated complete transcriptomes from Drosophila
internal organs and tissues into bacterial biomass, facili- wild-type adult flies infected by the insect pathogenic
tated by a wide range of toxins, virulence factors and nematodes Heterorhabditis and their mutualistic bacteria
hydrolytic enzymes produced by the bacteria [11, 12]. Photorhabdus, separately or together. We examined gene
For two to three generations the developing nema- transcription for two time-points, 12 and 30 h post-
todes feed on the bacterial biomass until the insect infection with the pathogens (Fig. 1). These time-points
carcass is consumed, whereupon adult development is correspond to the initial entry and spread of the patho-
suppressed and the IJ stage accumulates. These non- gens in the fly (12 h post-infection) and to the establish-
feeding IJ containing their mutualistic bacteria emerge mentofdisease(30hpost-infection)[27].Thenumbersof
into the soil to seek new hosts [13]. We and others sequencereadsmappedto80.28%oftheD.melanogaster
have previously shown that Heterorhabditis is a po- genome (Fig. 1a). Similarly, the high number of reads
tent pathogen of Drosophila, and have begun using sequenced had more than 90 % coverage of the D. mela-
the Drosophila-Heterorhabditis model system to under- nogaster genome (Additional file1: Figure S1). The quan-
stand the molecular interplay between insect immune titativereal-timeRT-PCR(qRT-PCR)analysisofrandomly
functionandnematodeparasiticstrategies[14–18]. selected genes (CG34040, CG64267, CG9468, CG11909,
Whole genome mRNA sequencing (RNA-Seq) tech- CG6524,CG17571,CG10374) using gene-specific primers
nologies have been a significant advance for high- (Additional file 1: Table S1) validated the RNA-Seq data
throughput transcriptome analyses, as they can gener- (Additional file 1: Figure S2). We found upregulation for
ate hundreds of millions reads in a single sequencing CG34040, CG64267, CG11909 and downregulation for
run [19, 20]. RNA-Seq is more sensitive, quantitative CG9468, CG6524, CG17571, CG10374 although the level
and efficient, and it has higher reproducibility com- oftranscriptiondetectedwashigherbyRNA-Seqwiththe
pared to previously used hybridization-based micro- exceptionofthegeneCG11909.
array techniques [21]. RNA-Seq has already produced The highest number of differentially expressed genes
exciting and novel information in the study of various was observed in flies infected by Photorhabdus at 30 h
diseases [22, 23]. This powerful tool is becoming in- post-infection(Fig.1b).Strikingly,wefoundthatthe vast
creasingly attractive for investigating the transcrip- majority of fly genes (82 %, 2555 genes) were downregu-
tional profiles in model and non-model organisms lated at 12 h post-infection with Photorhabdus, and
[24, 25]. Recent works have started to report the use similar numbers of genes were upregulated or downreg-
of RNA-Seq (Illumina or 454-pyrosequencing) for the ulated at 30 h after infection with the bacteria (2763 and
comprehensive understanding of the transcriptional 2845 genes, respectively). At 12 h post-infection with
regulation of genes that participate in pathogen viru- axenic Heterorhabditis, there were 1125 upregulated
lence and host innate immune processes [26]. genes (47 %) and 1238 downregulated genes (53 %).
Here we have infected Drosophila melanogaster adult Similarly, infection with symbiotic nematodes upregu-
flies with symbiotic Heterorhabditis (nematodes carrying lated 819 genes (48 %) and downregulated 871 genes
Photorhabdus), axenic Heterorhabditis (nematodes lack- (52 %). We also found that axenic nematodes at 30 h
ing Photorhabdus), and Photorhabdus bacteria alone and post-infection downregulated a higher number of fly
used RNA-Seq to analyze the transcriptional response genes (67 %, 2868 genes) compared to those downregu-
of flies to the pathogens, either separately or together. lated bysymbioticworms (51%,2002genes) (Fig. 1b).
Castilloetal.BMCGenomics (2015) 16:519 Page3of21
Fig.1InfectionofadultflieswithHeterorhabditisnematodesortheirPhotorhabdusbacteriaelicitsdistincttranscriptomicprofiles.aTranscriptome
summary(numberofreadsandpercentagemappedtotheD.melanogastergenome)fromfliesinfectedbyHeterorhabditisaxenicorsymbiotic
nematodes,orPhotorhabdusbacteriaat12and30hpost-infection.bDifferentialgenetranscription(upregulated/downregulatedgenes)inflies
at12hand30hpost-infectionwithHeterorhabditisaxenicorsymbioticnematodes,orPhotorhabdusbacteriaalone.cCUFFLINKSanalysisof
differentiallyexpressedtranscriptsbetweenthe12and30htime-pointsinfliesinfectedbyHeterorhabditisaxenicorsymbioticnematodes,or
Photorhabdusbacteriaalone.dVenndiagramsshowingthenumberofDrosophilagenesthataredifferentiallyexpressed(upregulatedor
downregulated)at12honlyorat30honlyoratbothtime-pointsafterinfectionwithHeterorhabditisaxenicorsymbioticnematodes,ortheir
Photorhabdusbacteriaalone.Expressionpatternsareindicated(UP/UP:geneupregulationatboth12and30h,DOWN/UP:genedownregulation
at12handupregulationat30h,DOWN/DOWN:genedownregulationatbothtime-points,UP/DOWN:geneupregulationat12hand
downregulationat30h)
Toidentifythe numbers ofdifferentiallyexpressed gene worms(618and154,respectively)andfewer (114) inflies
isoformsinducedbyeachpathogenateachtime-point,we infectedbyPhotorhabdus(Fig.1c).
performedCufflinksanalysis[28].Wefound273geneiso- To determine the number of genes that are transcrip-
forms upregulated by Photorhabdus, whereas infection by tionally regulated upon infection with Heterorhabditis
axenic or symbiotic nematodes upregulated fewer gene andPhotorhabdus,weperformedpairwise multiple com-
isoforms in the fly (131 and 52, respectively) (Fig. 1c). parison analyses. We found that the number of differen-
However, the largest number of downregulated gene iso- tially regulated genes at 12 and 30 h post-infection
forms was found in flies infected by symbiotic or axenic varied among the different types of infection (Fig. 1d).
Castilloetal.BMCGenomics (2015) 16:519 Page4of21
We focused on the group of genes that were regulated at but not in flies infected by axenic or symbiotic nema-
both time-points and observed: (i) several genes that were todes (Additional file 2: Dataset S1).
downregulated at 12 h post-infection and upregulated at PANTHER analysis pointed towards pathways that
30 h post-infection (28 genes in flies infected by axenic were not identified by KEGG analysis (Fig. 2d,e,f). Infec-
worms,482infliesinfectedbysymbioticwormsand97in tion of flies with the two pathogens for 30 h affected the
fliesinfectedbythebacteria);ii)ahighernumberofupreg- androgen/estrogen/progesterone biosynthesis pathway,
ulated genes in flies infected by axenic worms (528 genes) which includes genes that take part in lipid metabolism,
compared to those upregulated by symbiotic worms (265 steroid hormone metabolism, and cholesterol meta-
genes) and the bacteria alone (256 genes); iii) a smaller bolism. Infection with axenic or symbiotic nematodes
numberofgenesthatwereupregulatedat12handdown- caused the enrichment of genes involved in asparagine/
regulatedat30hbyinfectionwithaxenicnematodes,sym- aspartate biosynthesis. Infection with symbiotic worms
bioticnematodes,orPhotorhabdusalone(82,139and107 or Photorhabdus bacteria alone induced the enrichment
genes, respectively; (iv) a large number of genes that of genes involved in ubiquitin-proteosome degradation
remained transcriptionally downregulated upon infection of proteins, pyruvate metabolism and pentose-phosphate
withsymbioticworms,axenicwormsorthebacteriaalone pathway. Therefore pathway analyses reveal certain mo-
(963,577,and481,respectively).Theseresultssuggestthat lecular/biological signatures in the fly and provide hints
alargesetofgenesisdifferentiallyregulatedinDrosophila on the physiological events that take place during infec-
adult flies during the early and late stages of infection by tion withthenematodes andtheirassociatedbacteria.
Heterorhabditis nematodes and their mutualistic Photo-
rhabdusbacteria. HeterorhabditisandPhotorhabdusinfectionregulatesthe
transcriptionofawiderangeofprotein-codinggenesin
HeterorhabditisandPhotorhabdusinfectionleadsto theflygenes
changesinspecificmolecularpathwaysandbiological To identify the major protein families and biological pro-
activitiesinDrosophila cesses associated with the fly genes that are differentially
To identify the molecular pathways and biological activ- regulated by Heterorhabditis and Photorhabdus infection,
ities regulated by the nematodes and their bacteria, we we conducted gene ontology (GO) analysis [33] (Fig. 3a).
performed Database for Annotation, Visualization and At late stages of infection with the pathogens there was a
Integrated Discovery (DAVID) analysis by interrogating substantialincreaseinthenumberofprotein-codinggenes.
the Kyoto Encyclopedia of Genes and Genomes (KEGG) For example, we found a dramatic increase in the expres-
and Protein ANalysis THrough Evolutionary Relation- sionof hydrolasegenesuponinfectionwiththenematodes
ships (PANTHER) databases [29–32] (Fig. 2). At 12 h and their bacteria. At 30 h post-infection with the patho-
after infection with axenic or symbiotic nematodes, there gensthetop15proteincategoriesincludedhydrolases,nu-
was a significant change in the transcription of a large set cleic acid binding proteins, oxido-reductases, transferases,
of genes, which generated distinct pathway categories transporters, proteases, transcription factors, receptors, en-
(KEGG). For instance, we found that axenic and symbiotic zyme modulators, signaling molecules, cytoskeletal pro-
nematodeselicitedtheenrichmentofgenesinvolvedinme- teins, ligases, transfer/carrier proteins, calcium-binding
tabolismofxenobioticsbycytochromeP450andglycolysis, proteinsandkinases.
whilesymbioticnematodeselicitedtheenrichmentofgenes GO analysis also showed that infection with axenic
involved in lysosome function and apoptosis signaling worms strongly downregulated a large number of genes
(Additionalfile1:FigureS3;Additionalfile2:DatasetS1). associated with metabolic process at both 12 and 30 h
We also found that a large number of genes within cer- post-infection (356and700genes,respectively) (Fig. 3b).
tainpathways(suchaslimoneneandpinene,tyrosinedeg- However, the number of similar genes that were upregu-
radation, arginine and proline metabolism, biosynthesis of lated upon infection with axenic worms increased from
unsaturated fatty acids, drug metabolism, folate biosyn- 252 to 442. Genes involved in proteolysis and G-protein
thesis, glutathione metabolism and glycosylation) were coupled receptor (GPCR) signaling were strongly down-
transcriptionallyalteredfollowinginfectionofflieswiththe regulatedat30h,whileall7genesassociatedwithpepti-
nematodes and their bacteria (together or alone) for 30 h dyl hydroxylation were downregulated at 12 h post-
(Fig.2a,b,c).We further found downregulation of orni- infection. We further found that only 23 immunity-
thine decarboxylase related genes in all three types of related genes were upregulated at 30 h and 9 genes at
infections, which might suggest the induction of anti- 12h,while29genesinvolvedinchitin metabolicprocess
inflammatory responses in the fly (Additional file 2: were downregulated at30hpost-infection.
Dataset S1). There was also strong downregulation of GO analysis further revealed that infection of flies for
a large number of genes in the ubiquitin/proteasome 30 h with symbiotic nematodes downregulated several
degradation pathways in Photorhabdus infected flies, genes involved in proteolysis, response to stimulus,
Castilloetal.BMCGenomics (2015) 16:519 Page5of21
Fig.2InfectionofadultflieswithHeterorhabditisnematodesortheirPhotorhabdusbacteriainducesdiversephysiologicalresponses.Representative
KEGGpathwaycategoriesinfliesinfectedbyaHeterorhabditisaxenicnematodes,bHeterorhabditissymbioticnematodes,orcPhotorhabdus
bacteriaat30hpost-infection.Thenumberofgenesrepresentsthosethatwereonlyfoundassociatedwithaparticularpathway.Representative
PANTHERpathwaycategoriesinfliesinfectedbydHeterorhabditisaxenicnematodes,eHeterorhabditissymbioticnematodes,orfPhotorhabdus
bacteriaat30hpost-infection
chitin metabolic process, cell surface receptor signal with putative function in Notch signaling, peptidyl amino
transduction, and lipid metabolic process (Fig. 3c). In acidmodification,cellrecognition,andlymphglanddevel-
addition, Photorhabdus infection resulted in substantial opment(Fig.3d).At30h,therewasastrongdownregula-
increase in the number of upregulated and downregu- tion of genes involved in oxidative phosphorylation, lipid
lated genes involved in proteolysis, oxidation-reduction metabolic processes and ribosome function; the latter
and responses to stress at both 12 and 30 h time-points might be an indication of transcriptional repression
(Fig. 3d,e). At 12 h, we identified downregulated genes (Fig.3e).Wefurtherfoundupregulationofgenesinvolved
Castilloetal.BMCGenomics (2015) 16:519 Page6of21
Fig.3(Seelegendonnextpage.)
Castilloetal.BMCGenomics (2015) 16:519 Page7of21
(Seefigureonpreviouspage.)
Fig.3InfectionofadultflieswithHeterorhabditisorPhotorhabdustriggertheexpressionofdiverseproteins.aRepresentativeprotein-basedGene
Ontology(GO)groupsforgenesdifferentiallyinducedbyHeterorhabditisaxenicorsymbioticnematodes,orPhotorhabdusbacteriaaloneat12and
30hpost-infectionwiththepathogens.Eachbarrepresentsasubsetofthemostrepresentativenon-redundantupregulatedanddownregulated
genes.NumbersofupregulatedanddownregulatedgenesuponinfectionwithHeterorhabditisaxenic(b),symbiotic(c)nematodes,orPhotorhabdus
bacteria(d,e)at12hand30hpost-infectionwiththepathogens.Eachbarincludesgenesthatfallintothesamemolecularfunctioncategory.GO
analysiswasperformedusingthegloballistofdifferentiallyexpressedgenesforeachinfectiontypeandtime-point
in responses to hypoxia and insulin receptor signaling binding and cell adhesion [35], and the putative enzyme
pathway. These data show that infection of the fly with CG31002 that possesses glucuronosyltransferase activity.
Heterorhabditis nematodes and their Photorhabdus bac- Interestingly, we found no increased expression of anti-
teriacausessignificantchangesintheexpressionofalarge microbial peptide (AMP) genes (Fig. 4a); only Attacin C
number of protein-coding genes that are involved in key and Drosomycin 2 were upregulated upon infection with
biological processes in the fly. Some of the differentially symbiotic nematodes (Fig. 4c) and Photorhabdus bacteria
expressed genes might be important in the regulation of (Fig. 4e), respectively. At 30 h post-infection with axenic
immunefunctionagainstthepathogens. nematodes, we found increased expression of the AMP
genes Attacin and Drosocin, a Gram-Negative Binding
Thetranscriptomeoffliesinfectedbysymbiotic Protein(GNBP)-likegeneandseveralcuticle-relatedgenes
Heterorhabditisisacombinationofthetranscriptomes such as Tweedle, Cuticular protein 78E and two genes
fromfliesinfectedbyaxenicnematodesand with chitin-binding domains (CG7017 and CG6933)
Photorhabdus (Fig.4b).Wealsofoundseveralhighlyexpressedenzymes
To contrast the transcriptional profiles induced by the including a trypsin-like cysteine protease, a putative
nematodes and their bacteria at 12 h and 30 h post- AMP-dependent synthetase (CG4830), and a putative lip-
infection, we performedaquadrantplotanalysis to iden- ase(CG5665).
tify the genes that are differentially expressed at those Thetranscriptome of flies infected by symbiotic nema-
two time-points (Additional file 1: Figure S4). Wegener- todes was a combination of the transcriptomes obtained
ated three clusters containing differentially regulated genes from flies infected by axenic worms and the bacteria
at both time-points compared to uninfected treatments, alone. At 12 h post-infection, the gene Fat body protein
genes regulated at 12 h only, and those regulated at 30 h 1 was expressed at high levels, followed by the non-
only. Infection with axenic nematodes for 12 h strongly coding RNA Iab-8 that was previously shown to be
downregulated the genes CG34424 (5-formyltetrahydrofo- involved in the regulation of developmental processes
late cyclo-ligase), CG13071 (unknown), CG33264 (Or69a), [36] (Fig. 4c). We also observed upregulation of several
CG31748 (Gr36c), whereas infection with symbiotic nema- genes coding for ionotropic receptors and the odorant
todes mostly downregulated genes with unknown function receptor 59a. At 30 h post-infection with symbiotic
such as CG43184, CG7327, CG8960 and CG42755. Simi- worms we found increased expression of genes encond-
larly,Photorhabdusinfectioncauseddownregulationofsev- ing structural components of the cuticle (chitin), such as
eral unknown genes, such as CG13427, CG42367 (insect Lcp65Ag3, the Cuticular protein 67Fb and the adult cu-
cuticular protein) and CG13711. A complete list of the 25 ticular protein Accessory gland protein 54A1 (Fig. 4d).
moststronglydownregulatedgenesuponinfectionwiththe The gene Niemann-Pick type C-2d was also expressed at
twopathogensisshowninAdditionalfile1:FigureS5.We high levels. Interestingly, this gene codes for a sterol bind-
furtheridentifiedthe25moststronglyupregulatedgenesin ing protein with Immunoglobulin E-set and MD-2-related
fliesinfectedbyaxenicnematodes,symbioticnematodesor lipid recognition domains with a potential function in im-
thebacteriaonly(Additionalfile3:DatasetS2). mune recognition and defense. In addition, there were sig-
At 12 h post-infection with axenic worms, we detected natures of transcriptional regulation, as evidenced by the
increased expression ofseveralgenes in the Notch signal- increasedexpressionofthetranscriptionfactorE(spl)region
ing pathway (Fig. 4a). Among this group of genes, the transcript mdelta and the uncharacterized putative zinc-
negativeregulatorTwinofM4orBarbushowedthehigh- finger transcription factor CG14983. We further detected
est level of expression followed by a putative CCAT- theexpressionofCG16704,whichencodesaputativepro-
binding transcription factor, the gene Enhancer of split tein with Proteinase I2 and Kunitz protease inhibitor do-
mgamma, a basic Helix-Loop-Helix transcription factor mains that could be involved in coagulation response or
relatedtoMyc,BrotherofBeardedA,whichhasbeenpre- otherproteolyticcascades,severalsubunitsforH+ATPase
viouslyimplicatedintheflyimmuneresponseagainstbac- pumps, a putative transporter (CG14605), autophagy-
terial infection [34], Amalgam, which codes for an Ig-like specific geneAtg8b,Tweedle M, a putative memberofthe
C2-type domain-containing protein involved in antigen smallGTPasefamily(CG17819)andJonah65Aiiprotease.
Castilloetal.BMCGenomics (2015) 16:519 Page8of21
Fig.4HeterorhabditisandPhotorhabdusinducetheexpressionofdiversesubsetsofgenesinDrosophilaadults.The25moststronglyinduced
genesuponinfectionwitha,bHeterorhabditisaxenicnematodes,c,dHeterorhabditissymbioticnematodesande,fPhotorhabdusbacteriaat
12hand30hpost-infection.X-axisrepresentstherelativeLog-FoldChange(LFC)foreachgeneafternormalizationagainstuninfectedcontrols.
Allgeneshaveafold-changehigherthan2(LFC=0.58correspondsto2-fold-changedifference)
Photorhabdus induced a distinct transcriptional profile were upregulated at the early phase of infection with
in the fly compared to the nematodes. At 12 h post- Photorhabdus included genes coding for cuticular
infection we identified a large number of genes with un- proteins (Adult cuticle protein 1 and Adult cuticle
known function or identifiable protein domains (Fig. 4e). protein 65Aa) and Jonah proteases. We also found a
For example, the gene CG12998 with unknown function CD36 antigen domain-containing gene (CG2736) and
was expressed at the highest level, followed by CG18179 the microRNA (miRNA) Mir-2494 stem loop. At 30 h
that codes for a putative peptidase. Other genes that post-infection with the bacteria, there was increased
Castilloetal.BMCGenomics (2015) 16:519 Page9of21
expressionofdetoxificationgenes(CytochromeP450-4e3), vs. Symbiotic nematode infections and Axenic Hete-
the cytokine Unpaired-3 (component of the Janus kinase/ rorhabditis vs. Photorhabdus infections (Fig. 5b). We
signal transducers and activators of transcription or JAK/ observed contrasting gene expression levels between
STATpathway)[37],variousTweedlegenesandtwoOsiris the two comparisons. We also found that although
genes (Fig. 4e). These results indicate that infection with most of the highly regulated genes have an unknown
HeterorhabditisnematodesandPhotorhabdusbacteriadif- function, gene Lcp65Ag1 that is involved in the struc-
ferentially regulates several Drosophila genes, many of ture of chitin-based cuticle was predominantly upreg-
whichhaveanuncharacterizedfunctioninthefly. ulated in flies infected by Heterorhabditis nematodes
[44], whereas gene Niemann-Pick type C-1b that is
HeterorhabditisnematodesandPhotorhabdusbacteria involved in central nervous system development was
differentiallyregulatesignaturegenesinDrosophilaflies highly upregulated by Photorhabdus infection [45].
We used the 12 h and 30 h time-points as replicates to Finally,wecontrastedtheexpressionofthosegenesthat
perform General Linear Model (GLM) analysis of genes were significantly affected when comparing Symbiotic
that are differentially expressed upon infection with Het- Heterorhabditis vs. Photorhabdus infections and Axenic
erorhabditis nematodes and their mutualistic Photorhab- Heterorhabditis vs. Photorhabdus infections (Fig. 5c).
dus bacteria, separately or together (Fig. 5 and Additional Thesecomparisonsprovideinsightsintotheexpressionof
file4:Dataset S3).For this, we first compared normalized the Drosophila genes that are mainly upregulated upon
genes(againstuninfectedcontrols)thatweredifferentially Photorhabdus infection, since downregulation of genes in
expressed upon infection with the pathogens. We then the comparison Axenic vs. Symbiotic Heterorhabditis was
builtmodelstoidentifythe group of genes whose expres- less than 2-fold. Again we found that AMP genes (Atta-
sion levels were significantly affected after comparing the cin-A, Cecropin A1, Cecropin B, Diptericin B and Listeri-
different infection treatments (Axenic Heterorhabditis vs. cin) were upregulated more than 2-fold by Photorhabdus
Symbiotic Heterorhabditis, Symbiotic Heterorhabditis vs. infection. We also found that the JAK/STAT pathway
Photorhabdus, and Axenic Heterorhabditis vs. Photorhab- cytokines Unpaired-2 and Unpaired-3 were also signifi-
dus). Therefore we exclusively selected those genes that cantly upregulated by the bacteria as well as the secreted
showeda significant(2-fold)changeinexpression(upreg- recognition protein PGRP-SD. Other upregulated genes
ulationordownregulation). includedSerineProteaseImmuneResponseIntegratorthat
We first compared the expression of genes that were af- is involved in the response against bacterial infections
fected when comparing all three types of infections using [46], the putative regulators of proteolysis Serpin 88Eb
the two time-points as replicates (Fig. 5a). We found that and Peptidase M13, the MAP kinase P38c that has been
most genes associated with Heterorhabditis nematodes shown to function in the intestine to regulate lipid me-
(Axenic vs. Symbiotic) were significantly downregulated tabolism and immune homeostasis [47], Sugarbabe that
whereasmostgenesassociatedwithPhotorhabdus(Symbi- encodes a zinc finger protein responsible for the regula-
otic Heterorhabditis vs. Photorhabdus and Axenic Hetero- tion ofinsulin geneexpression inthe neurosecretory cells
rhabditis vs. Photorhabdus) were significantly upregulated. [48],GlutathioneSTransferaseD2thatregulatesdetoxifi-
WeobservedthatseveralgenesencodingAMP(Attacin-B, cation [49],Yellow-F that encodes an enzyme responsible
Attacin-D, Attacin-C, Cecropin-C, Diptericin, Drosocin, for catalyzing the conversion of dopachrome into 5,6-
Drosomycin-2, Metchnikowin) as well as Lysozyme X were dihydroxyindole in the melanization pathway [50], and
downregulated in nematode-infected flies and upregulated Gustatory Receptor 94a that is a candidate taste receptor
by the bacteria. Other genes including the stress-related inDrosophila[51].DownregulatedgenesbyPhotorhabdus
genesTurandot (Tot), elevated during infection (edin), the includedJonah99Ciwithputativeendopeptidaseactivity,
Peptidoglycan Recognition Protein gene PGRP-SB1, a Lip3withputativelipaseactivity, Artichokethatencodesa
galactose-specific C-type lectin (Lectin-37 Da), and the leucine-rich repeat extracellular matrix protein required
DNAbindingtranscriptionfactorSox21a,whichhavebeen fornormalmorphogenesisandfunctionofciliatedsensilla
shown previously to participateintheDrosophila immune in Drosophila [52], and Magro that encodes a lipase A
response[38–43],werealsostronglyupregulatedbyPhoto- homolog that is secreted from the anterior gut into
rhabdusbutnotbyHeterorhabditisinfection.Interestingly, the intestinal lumen to digest dietary triacylglycerol and
two genes that are involved in metabolic processes, an hydrolyzecholesterolesters[53].
AMP-dependent synthetase/ligase and the putative ATP-
binding Pugilist, were upregulated in response to Hetero-
rhabditis infectionand downregulateduponPhotorhabdus InfectionwithHeterorhabditisandPhotorhabdusaltersthe
infection. transcriptionofkeyimmunegenesinthefly
Wethencomparedtheexpressionoffilteredgenesthat To identify which immune-related pathways in the fly
were differentially regulated when comparing Axenic are regulated upon infection with Heterorhabditis and
Castilloetal.BMCGenomics (2015) 16:519 Page10of21
Fig.5DifferentialgeneexpressionanalysisusingDESeqandGLManalysistocomparethedifferentinfectiontypes.aFilteredlistofgenesthatare
commonbetweenallthreeinfectiontypes(p<0.05);bFilteredlistofgenesthatarecommonbetweentheAxenicHeterorhabditisvs.Symbiotic
HeterorhabditisandAxenicHeterorhabditisvs.Photorhabduscomparisons;cFilteredlistofgenesthatarecommonbetweentheSymbioticHeterorhabditis
vs.PhotorhabdusandAxenicHeterorhabditisvs.Photorhabduscomparisons.ThefigurecontainsDrosophilageneswithsignificantlyalteredexpressionupon
infectionofadultflieswiththenematodesandtheirassociatedbacteria(separatelyortogether),andtheircorrespondingadjustedp-valuesforthetwo
modelsused:onetodeterminethecommongenesbetweenHeterorhabditis(Axenic/Symbiotic)infectionsandPhotorhabdusinfections,andasecondto
determinethecommongenesbetweenAxenicHeterorhabditisinfectionandPhotorhabdusinfectionafteradjustingforthetwotime-points.Selectedgenes
includedthosethatappearedinallthreecomparisons.Red:LFC≥2-folddownregulation;Yellow:LFC≤2-folddownregulationandGreen:LFC≥2-fold
upregulation