Transcriptome of the Plant Virus Vector Graminella nigrifrons, and the Molecular Interactions of Maize fine streak rhabdovirus Transmission

Background Leafhoppers (Hemiptera: Cicadellidae) are plant-phloem feeders that are known for their ability to vector plant pathogens. The black-faced leafhopper (Graminella nigrifrons) has been identified as the only known vector for the Maize fine streak virus (MFSV), an emerging plant pathogen in the Rhabdoviridae. Within G. nigrifrons populations, individuals can be experimentally separated into three classes based on their capacity for viral transmission: transmitters, acquirers and non-acquirers. Understanding the molecular interactions between vector and virus can reveal important insights in virus immune defense and vector transmission. Results RNA sequencing (RNA-Seq) was performed to characterize the transcriptome of G. nigrifrons. A total of 38,240 ESTs of a minimum 100 bp were generated from two separate cDNA libraries consisting of virus transmitters and acquirers. More than 60% of known D. melanogaster, A. gambiae, T. castaneum immune response genes mapped to our G. nigrifrons EST database. Real time quantitative PCR (RT-qPCR) showed significant down-regulation of three genes for peptidoglycan recognition proteins (PGRP – SB1, SD, and LC) in G. nigrifrons transmitters versus control leafhoppers. Conclusions Our study is the first to characterize the transcriptome of a leafhopper vector species. Significant sequence similarity in immune defense genes existed between G. nigrifrons and other well characterized insects. The down-regulation of PGRPs in MFSV transmitters suggested a possible role in rhabdovirus transmission. The results provide a framework for future studies aimed at elucidating the molecular mechanisms of plant virus vector competence.


Introduction
Hemipteran insects such as aphids, whiteflies, planthoppers and leafhoppers are arguably the most important vectors of plantinfecting viruses. These insects have specialized mouthparts suitable for tissue specific feeding (often the phloem), and wide host ranges that provide ample opportunity for virus transmission [1,2]. Most insect vectors of plant viruses have significant associations with humans and agroecosystems. Recent rapid changes in these environments have increased crop exposure to viruses and vectors or altered evolutionary, ecological or genetic interactions leading to enhanced transmission [3]. A lack of understanding these factors, including the molecular mechanisms of virus transmission by vectors, reduces our ability to assess and manage risks posed by plant virus vectors, particularly for emerging diseases whose full impacts are not yet realized.
Graminella nigrifrons is one of the most common and abundant leafhoppers in the eastern half of the U.S., presently found in 35 states from southern Maine to Florida [4]. It has a wide plant host range, including oats (Avena sativa L.), maize (Zea mays L.), perennial rye grass (Lolium perenne L.) and johnsongrass (Sorghum halepense L.) [5]. G. nigrifrons is a natural and experimental vector of several pathogens (e.g. maize bushy stunt phytoplasma and corn stunt spiroplasma) as well as several types of viruses, including rhabdoviruses [6][7][8][9][10].
G. nigrifrons has recently been identified as a vector of the emerging Rhabdovirus Maize fine streak virus (MFSV) [7,11], which is a newly described member of Nucleorhabdovirus, first detected in Georgia, US in 1999 [7]. In maize, this emerging virus causes symptoms of dwarfing and fine chlorotic steaks along intermediate and small veins. Consistent with most rhabdoviruses, MFSV has high vector species specificity as G. nigrifrons is the only reported insect vector [7,11,12]. G. nigrifrons transmits MFSV in a persistent propagative manner, but, vector competence for MFSV is not consistent within G. nigrifrons populations. Individual insects can be categorized into three groups according to their acquisition or transmission capabilities after exposure to MFSV infected plants. 'Transmitters' acquire and transmit virus (i.e., they are vectors); 'acquirers' acquire, but cannot transmit virus; and non-acquirers do not acquire virus after feeding on infected plants [7,11]. Mechanisms of these differences are unknown, but it is hypothesized that barriers to virus replication and movement within the insect might be the key factors limiting vector competence among G. nigrifrons individuals [13,14]. While biological parameters associated with MFSV and other pathogen transmission by G. nigrirons are defined, the absence of molecular resources for this vector prevents further investigations into insect vector competence, the molecular basis of rhabodvirus transmission, and comparative vector genomics. While a few molecular investigations exist for planthoppers [15,16], data for leafhoppers is virtually non-existent.
Rhabdoviruses can infect both plants and animals (including mammals) and consist of a nucleocapsid containing a negative single-stranded RNA genome encapsidated into a lipid bilayer membrane. There are seven recognized Rhabdoviridae genera, but only viruses within the Nucleorhabdovirus and Cytorhabdovirus infect plants [17,18]. In contrast to most plant viruses, the rhabdoviruses also infect their insect vectors, they must, therefore, replicate within the insect and cross several molecular and physical barriers prior to being transmitted to additional host plant [14]. These barriers may include digestive enzymes, peritrophic lining, improper epithelial cell receptors for virus attachment, programmed cell death, virus induced RNA interference (RNAi) or an active insect immune response [14,19]. A midgut barrier to virus infection of poor vectors has been defined for some rhabdoviruses, but other barriers are not as well understood [14,18].
In other systems, the impact of several insect responses on vector competence has begun to be defined, including insect innate immunity, a highly conserved, essential defense against microbial infection, such as bacteria, fungi, nematodes and viruses [20]. Non-self molecules are recognized and subsequently trigger cell-mediated phagocytosis, encapsulation and/or humoral responses. The humoral response is activated by the binding of peptidoglycan recognition proteins (PGRPs) to pathogen associated molecular patterns (PAMPs), which then regulate antimicrobial peptide (AMPs) production through Toll or immune deficiency (IMD) signaling cascades [21]. Two classes of PGRPs are described: the short PGRPs (PGRP-S) are small extracellular proteins without transmembrane domains present in the hemolymph and cuticle; and, long PGRPs (PGRP-L) are larger proteins that encode transmembrane domains [22]. The responses of insect PGRPs to bacterial and fungal infection have been well described, and several recent studies revealed the involvement of the Toll pathway and PGRPs in D. melanogaster defense against viral infection [19,23]. The number and the type of PGRPs vary among insects. For example, D. melanogaster has 12 PGRP genes [22] and Anopheles gambiae has 7 [24], but the Hemipteran pea aphid Acyrthosiphon pisum lacks PGRPs due to the co-evolution with its endosymbionts [25]. The role of PGRPs in leafhoppers and planthoppers, the primary vectors of plant viruses, is not well understood.
RNA interference (RNAi) is another key pathway protecting eukaryotic organisms from virus infection [26,27]. Genes, proteins and pathways important for RNAi have been characterized in D. melanogaster, but genes involved in these pathways have just begun to be described in insect vectors of plant viruses including Peregrinus maidis [28], and A. pisum [29]. No information about RNAi pathway genes is available for leafhoppers.
To expand leafhopper molecular resources, we constructed cDNA libraries using RNA from G. nigrifrons transmitters and acquirers, and sequenced the transcripts using RNA-seq. The resulting EST database provides the first transcriptome information on a leafhopper, and contains genes potentially involved in anti-pathogen defense. In addition, we used RT-qPCR to examine the expression of seven genes putatively involved in insect immune response for leafhoppers known to transmit MFSV and leafhoppers not previously exposed to MFSV.

Insect rearing and virus maintenance
G. nigrifrons was maintained on maize (Zea mays L.) hybrid Early Sunglow in growth chambers, and MFSV was maintained by serial transfer to maize at OARDC as previously described [11]. For virus maintenance, 300 G. nigrifrons individuals were reared on MFSV infected maize for three weeks then transferred to healthy maize seedlings. Inoculated plants with MFSV symptoms were subsequently used as source plants.

Differentiating non-acquirers, acquirers, and transmitters
Two hundred G. nigrifrons adults (100 males and 100 females) were collected, placed in a single cage, and allowed to mate and oviposit on MFSV infected maize seedlings for two days. After 14 days, the (F 1 ) nymphs were observed, and this was denoted as day 0. The F 1 nymphs were reared on MFSV infected maize seedlings for 21 days, then F 1 adults were individually transferred to 4-day old healthy maize plants within a tube. After one week, F 1 leafhoppers were collected, labeled according to the maize test plant, and stored at 280uC. Plants were moved to a growth chamber for three weeks for MFSV symptom development [11]. An insect was designated as a transmitter if MFSV symptoms developed on the test plant it fed on. For non-transmitter insects, RT-PCR was performed on individuals using the Access RT-PCR System (Promega, Madison, WI) following the manufacturer's protocol. Primer pairs 361F (59-GTGCAGAATTGCCCTATCC-39)/917R (59-TCGAGGCAATTCCTGTATC-39) and 5335F (59-CTCCCATTATCATAGATAAAG-39)/6360R (59-TATATG-CAATTCTGATTCCTC-39) were used to amplify a 1120 bp and 1030 bp fragments of the viral N and G genes, respectively, based on MFSV genome sequence [12]. Reverse transcription was performed at 45uC for 45 min and followed by 2 min at 94uC. PCR included 40 cycles of 94uC for 30 s, annealing at 54uC (N gene) or 52uC (G gene) for 30 s, and extension at 68uC for 120 s. An insect was designated as an 'acquirer' if RT-PCR indicated the presence of MFSV, but no transmission of the virus to the test plant. 'Non-acquirers' were those insects for which MFSV was not detected by RT-PCR, and no symptoms developed on test plants.

RNA isolation and cDNA library construction
Total RNA from individual G. nigrifrons transmitters and acquirers was isolated with Trizol (Invitrogen, Grand Island, NY) following the manufacturer's protocol. The concentration and the quality of RNA were analyzed using a Nanodrop 2000c spectrophotometer and Agilent 2100 Bioanalyzer. RNA (10 mg) was pooled from eight individuals from each class and used to construct two cDNA libraries following the mRNA sequencing sample preparation guide (Illumina, San Diego, CA). Paired-end DNA sequencing was done in two lanes (one per library) on an Illumina GA-II following manufacturer's protocol.

Sequence assembly and annotation
The 76-bp paired-end Illumina reads from the acquirer (32,548,016 reads) and transmitter (30,541,892 reads) libraries were combined for de novo assembly. Low-quality ($80% of the read with the Phred score of less than 20) and low complexity (.80% of the read with single-nucleotide, di-nucleotide, or trinucleotide repeats) were removed. The processed reads were then assembled using a combination of the Velvet (ver. 1.2.01) [30] and Oases (ver. 0.2.06; http://www.ebi.ac.uk/zerbino/oases/) programs with the k-mer lengths of 41,43,45,47,49,51,53, and 55. The resulting assembled sequences and singletons were combined, processed to remove duplicates using a custom Perl Script [31], and further assembled after examining the overlapped regions identified by Vmatch (ver. z.z) [32]. Contigs were then further assembled using Phrap program (version: 0.020425.c) [33] to obtain the final transcriptome of sequences .100 bp. Sequences have been submitted to the short read archive at NCBI GenBank under accession number SRP013390.3.
Functional annotation of the G. nigrifrons transcriptome was performed by searching for analogous sequences in the Swiss-Prot database (http://www.ebi.ac.uk/uniprot) using an E-value cut-off of 10 24 . Hierarchical functional categorization on gene ontologies (GO terms) was done using BLAST2GO (http://www.blast2go. de) [34]. BLAST2GO was also used to identify genes represented in Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways (http://www.kegg.com/kegg/kegg1a.html).

Validation of reference genes for gene expression studies
Six candidate housekeeping genes were selected: alpha tubulin (a -TUB) (GnigEST017131), elongation factor 1-alpha (EF-1a) (Gni-gEST008963), glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (Gni-gEST000002), succinate dehydrogenase (SDHA) (GnigEST008732), ribosomal protein L3 (RPL3) (GnigEST000330), ribosomal protein S13 (RPS13) (GnigEST004586). Six treatments were used including MFSV-transmitters, non-acquirers and leafhoppers raised on healthy plants, separated in groups of males and females. Total RNA was extracted from individual G. nigrifrons adults using Trizol (Invitrogen). DNA was removed from the RNA samples with the TURBO DNA-free kit (Ambion, Grand Island, NY). The concentration and quality of the RNA were verified as outlined above. Total RNA (1mg) from individual insects was used for cDNA synthesis with the SuperScript III First-Strand Synthesis System (Invitrogen, Grand Island, NY). The cDNA was used as template for RT-qPCR using the iQ SYBR Green Mix (Bio-Rad, Hercules, CA), following the manufacturer's recommendations and primers designed using Beacon Designer 7.0 (BioRad, Hercules, CA). PCR was performed with 5ml of SYBR Green Mix, 2 ml of cDNA template (20ng/ml) and 5 pmole of each primer. Target genes were amplified at 95uC for 3 min followed by 40 cycles of 95uC for 10 s and 60uC for 30 s. Nuclease-free water was used for negative control reactions. PCR efficiency (E) was calculated by the equation E = 10 [21/slope] [37]. A standard curve was constructed using serial dilutions of pooled individual cDNAs and used to determine the relative expression values for these six reference genes (User Bulletin #2: ABI Prism 7700 Sequence Detection System vide supra) (http://www3. appliedbiosystems.com/). The software geNorm was used to determine the most stable reference gene among the six tested [38].
Expression of seven genes with known function in insect innate immunity was evaluated in transmitters and leafhoppers raised on healthy plants (i.e. never exposed to MFSV infected plants, used as control) using RT-qPCR: acetylcholine receptor subunit alpha-L1 (AChR) (GnigEST002867), autophagy protein 5 (ATG 5) (GnigEST012478), defensin (GnigEST011432), peptidoglycan recognition protein SB1 (PGRP-SB1) (GnigEST015027), PGRP-SD (GnigEST009213), PGRP-LC (GnigEST006324), tripeptidyl peptidase II (TPP II) (Gni-gEST021246). RNA isolation and RT-qPCR protocols were carried out as outlined above with primers listed in Table S1. The expression profiles of the seven genes were normalized to the internal control RPS13. Nuclease-free water was used as the negative control reactions. For each treatment, three biological replicates composed of RNA isolated from 10 adult leafhoppers were analyzed. The relative accumulation of transcripts in MFSV transmitters and control leafhoppers were determined using the comparative C T method [39]. A T-test was used to compare mean 2 2DCt values.

Sequence assembly and annotation
A total of 38,240 good quality ESTs of a minimum 100 bp were generated for the G. nigrifrons transcriptome. Approximately 34% of these transcripts (n = 13,036) mapped to the Swiss-Prot database (E,10 24 ) based on deduced amino acid similarity. Of these, 177 homologs were of prokaryotic origin and therefore removed from further analysis. On a more conserved level (E-value ,10 2180 ), 1,070 G. nigrifrons ESTs have high similarity with genes in the Swiss-Prot database (listed in Table S2). Approximately one-third of these highly conserved genes (n = 355) had highest similarity to a D. melanogaster gene. To guide biological interpretation, we used BLAST2GO to examine the associations between gene (GO) and enzyme (EC) ontologies assigned to the G. nigrifrons ESTs. We identified 4,488 ESTs (12%) that mapped to GO or EC terms ( Figure S1). Of these, 2,634 G. nigrifrons ESTs were classified as having function in cellular processes, 2,592 as having binding function, and 2604 as cell parts. KEGG-based pathway analysis using BLAST2GO from ortholog cellular pathways indicated that 882 ESTs could be putatively assigned to one or more of 126 KEGG pathways (Table S3). Thus, our initial functional analysis of the G. nigrifrons transcriptome indicates significant similarities with those of previously sequenced organisms, particularly D. melanogaster. However, fewer than half of the G. nigrifrons ESTs had significant similarity to genes in SWISS-PROT database, and less than 10% could be classified by GO term or KEGG pathway.

Comparison of G. nigrifrons and other invertebrate ESTs
Pair-wise comparison of the G. nigrifrons ESTs with cDNA databases for seven well-characterized invertebrate genomes and one insect dbEST collection indicated that ca. 37% (n = 14,259) of the G. nigrifrons ESTs had a significant match to at least one of the insect databases (E,10 210 ). When significant G. nigrifrons matches to the same predicted or known ortholog protein sequence in different species were collapsed into a single observation, there were 9,635 transcripts with at least one hit to an invertebrate database (,25% of the total ESTs). Partitioning of the significant matches among the eight invertebrate databases indicated D. melanogaster (n = 7,303) and T. castaneum (n = 7,028) had the greatest number of matches, whereas the most distantly related species, C. elegans, had the fewest (n = 4,777). The number of matches ranged between 6,309 and 6,907 for the remaining five invertebrate comparisons. Differences in the number of matches may be due to genomic evolution, or may more likely reflect different stages of transcriptomic characterization and curation.
For the putative or known ortholog transcripts, the number of G. nigrifrons ESTs that match a sequence in only one of the eight comparison invertebrate species was determined. The number of significant ortholog matches as well as the number of transcripts exclusive to one invertebrate for all eight pair-wise comparisons are shown in figure 1. Roughly half of the unique matches exclusive to one invertebrate (161 of 332) were attributed to N. vitripennis or the hemipteran A. pisum. The relatively low percentage (3%) of unique matches to the most closely related species in the comparisons, the corn planthopper P. maidis, is likely due to the small number of P. maidis ESTs (10,636) compared to the transcriptomes of the other seven query species.

Immune response ESTs in G. nigrifrons
The transcriptional responses of pathogen infection in insect human disease vectors are well described [40][41][42][43][44][45][46]. These studies suggest that innate immunity genes can have different functions in different genetic backgrounds, between closely related insect vector species, and even between different populations of the same species [47]. However, little is known about the transcriptome response in insect plant disease vectors due to pathogen invasion. To identify ESTs that play a putative role in the leafhopper immune response, we compared our dataset against .300 A. gambiae, D. melanogaster, and T. castaneum transcripts with known function in the insect immune response (including several transcript variants of some genes). The 25 G. nigrifrons ESTs with highest similarity to known immunity genes of A. gambiae, D. melanogaster, and T. castaneum are shown in Table 1.
A total of 194 G. nigrifrons ESTs were predicted to be functional in the immune response (Table S4) such as Toll and immune deficiency (IMD) pathways, Jun N-terminal kinase (JNK), and Janus Kinase/signal transducer of activator of transcription (JAK/ STAT). In addition, 10 ESTs with the E#10 213 , putatively function as pattern recognition proteins (PRPs) including different gram-negative bacteria binding proteins (GNBPs), C-type lectin and scavenger receptors. Of particular interest in this group of PRPs are PGRPs, which are the proteins responsible for sensing and binding of non-self molecules and which activate downstream immune responses. Since the Toll pathway was reported to play an important role in Drosophila X virus (DXV) defense [23], ESTs encoding putative key proteins involved in this pathway such as Spaetzle, protein Toll as well as its partners, Pelle and MyD88, Cactus, and Dorsal-related immunity factor (Dif) were identified. Other essential proteins function in IMD pathway such as DREDD and Relish, their corresponding ESTs were also present in G. nigrifrons. Although there are eight types of AMPs identified in D. melanogaster, there was only one putative AMP gene, defensin, identified in G. nigrifrons ESTs. The relatively large percentage of highly similar transcripts is expected, given the high degree of conservation of immune signaling pathways across diverse insect and mammalian taxa [48,49].

G. nigrifrons ESTs with similarity to RNAi pathway genes
RNAi is a highly conserved gene silencing process triggered by double-stranded RNAs and guided by small interfering RNA (siRNA) that involves post-transcriptional gene silencing (PTGS), which is essential for virus defense and development in insects [50,51]. Thirty-two Drosophila genes have been implicated in RNAi/PTGS. G. nigrifrons ESTs homologous to RNAi/PTGS  (Table 2). Significant matches for some of the D. melanogaster RNAi/PTGS genes (e.g. FBgn0262447, FBgn0262432, FBgn0262391) were not identified, likely due to their small size (,100 bp).

Validation and expression profiles of G. nigrifrons candidate genes
Since reference genes for assaying G. nigrifrons gene expression have not been described, we examined expression of six genes previously selected for reference gene validation in other insects using RT-qPCR [52]. Primer pairs for all six genes had a R 2 .0.99 and primer efficiency (E values) between 1.8 and 2.2 (Table S5). Stability analysis revealed RPS13 and a -TUB were the most stably expressed genes ( Figure S2), and RPS13 was chosen based on comparisons with other insects [53].
Gene expression patterns for seven selected genes implicated in insect innate immunity were tested using RT-qPCR, including three PGRPs (two representing short class and one representing long class), as well as AChR, ATG 5, TPP-II and defensin (the only AMP EST in our database) [19,28,54]. RNAs were isolated from G. nigrifrons individuals that were either experimentally determined to be MFSV transmitters or control leafhoppers. RNA used for RT-qPCR had an A 260 /A 280 between 1.8 and 2, and the PCR efficiencies for the candidate gene primers were between 1.8 and 2.2 (Table S1). The fold difference in expression for transmitters versus control leafhoppers ranged from 24.0 to 1.4 ( Figure 2). No significant differences in ATG5, AChR, TPP-II or defensin expression were detected. However, expression of the PGRP genes was significantly lower in transmitters compared to control leafhoppers.
Previous studies suggested roles for ATG5, AChR, TPP-II or defensin in the innate immune response. A role for autophagy in controlling vesicular stomatitis virus (VSV) replication was indicated in D. melanogaster [55]. TPP-II, a multi-function peptidase, was reported to be involved in protein turnover and Table 1. G. nigrifrons ESTs with highest sequence similarity to known immunity genes of A. gambiae, and D. melanogaster. immune response through processing antigen epitopes [56,57]. AChR was suggested as a receptor for rabies virus (Rhabdoviridae) [58]. Defensin is induced during bacterial infection after activation of the Toll pathway by the recognition and binding of PAMPs to PGRP-SA or SD [48,59,60]. However no significant expression difference between transmitters and control leafhoppers was detected for any of these four genes. A similar lack of an effect of MMV infection was found in P. maidis for ATG3 and TPP-II [28], and for defensin in Sigma virus (SIGMAV)-infected and noninfected Drosophila [19]. Therefore, although these genes have roles in innate insect immunity, their expression does not seem to  [19], and our data shows a potential interaction between putative G. nigrifrons PGRPs and rhabdovirus infection. All three putative-PGRPs tested (PGRP-SB1, PGRP-SD and PGRP-LC) showed significantly lower expression in MFSV transmitters compared to control leafhoppers, and the fold changes ranged from 22.9 to 24.0 ( Figure 2). Pathogens, such as bacteria and fungi, are recognized by PGRPs, which subsequently trigger downstream AMPs production through Toll or IMD immune pathway [22]. In Drosophila, infection with the rhabdovirus SIGMAV also alters expression of PGRP-SB1 and PGRP-SD; however in this case, increased transcript levels were found with these genes in infected Drosopila compared to un-infected [19,23]. Several non-mutually exclusive factors may explain the different expression profiles of these three putative PGRPs in the G. nigrifrons/MFSV and D. melanogaste/SIGMAV systems. SIGMAV are known insect pathogens, whereas the pathogenicity of MFSV in G. nigrifrons has not been determined. For example, SIGMAV, which is also a member of Rhabdoviridae, is a natural viral insect pathogen widespread in D. melanogaster populations, and is vertically transmitted [61]. Alternatively, MFSV is transmitted only through feeding on infected plant hosts; vertical transmission has not been observed in G. nigrifrons [7,11]. The difference in virus acquisition and pathogenicity may elicit different immune responses.
Additionally, we hypothesize that MFSV might actively manipulate vector immune response in terms of reduced expression of these three PGRPs in order to successfully replicate within G. nigrifrons transmitters and acquirers. Virus manipulation has been observed with the Togavirus Semliki Forest Virus (SFV) within infected cell lines of the mosquito Aedes albopictus; in this case, activation of immune defense pathways was suppressed [62]. Manipulation may be through direct interactions with the insect vector, or by altering physiological processes of the plant hosts such that an elevated plant defense response impacts insect defenses which consequently favors virus replication. Although this hypothesis is untested with G. nigrifrons-MFSV, further work aimed at describing the expression profiles of other PGRPs among all three G. nigrifrons classes (transmitters, acquirers, non-acquirers) will help understand the interactions between PGRPs and MFSV proteins that lead to virus transmission.
Our study represents the first transcriptome characterization of a leafhopper species, which includes 38,240 transcripts. Significant sequence similarity existed in immune defense genes existed between G. nigrifrons and other well-characterized insects. Additionally, we have identified several putative components of the RNAi pathway, which will enable future development of functional evaluation of the role of this important pathway in transmission of MFSV. The down-regulation of PGRPs in MFSV transmitters suggests a possible interaction with rhabdovirus transmission by vectors, although additional research is required to define the mechanism. The results presented expand molecular characterization of plant virus vectors and will help further understand the mechanisms of plant virus vector competence. Figure S1 Gene ontology (GO) terms for G. nigrifrons EST (contigs and singletons). The pie charts were generated Figure 2. Expression profiles of genes involved in G. nigrifrons immune response detected by RT-qPCR. Each bar represents the fold change of G. nigrifrons transmitters versus control leafhoppers. The gene expression level was nomalized to ribosomal protein S13 (RPS13), and the expression level of control leafhoppers was used as calibrator to calculate the fold change. Fold change was calculated based on a mean of three biological replicates concluding 30 individuals. Bars above the X axis show the genes expressed slightly higher in transmitters, while the bars below are genes down regulated in transmitters. *P,0.05; **P,0.01. doi:10.1371/journal.pone.0040613.g002 based on A. Biological process; B. Molecular function; C. Cellular component. (PDF) Figure S2 Comparison of reference genes for G. nigrifrons using geNorm. Genes with lower average expression stability M are more stable among all treatments. a-TUB, alpha tubulin; EF-1a, elongation factor 1-alpha; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; SDHA, succinate dehydrogenase; RPL3, ribosomal protein L3; RPS13, ribosomal protein S13. (PDF)

Supporting Information
Table S1 Primer sequences and efficiencies of candidate genes evaluated for differential expression among MFSV transmitters and control G. nigrifrons (RT-qPCR). (DOCX)