Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Reference genes selection for quantitative gene expression studies in tea green leafhoppers, Empoasca onukii Matsuda

  • Yongchen Yu,

    Roles Conceptualization, Data curation, Formal analysis, Methodology, Writing – original draft

    Affiliations Tea Research Institute, Chinese Academy of Agricultural Sciences, Hangzhou, Zhejiang, China, Key Laboratory of Tea Biology and Resources Utilization, Ministry of Agriculture, Hangzhou, Zhejiang, China, College of Horticulture, Northwest A&F University, Yangling, Shaanxi, China

    ⨯
  • Jin Zhang,

    Roles Data curation, Methodology, Supervision

    Affiliations Tea Research Institute, Chinese Academy of Agricultural Sciences, Hangzhou, Zhejiang, China, Key Laboratory of Tea Biology and Resources Utilization, Ministry of Agriculture, Hangzhou, Zhejiang, China

    ⨯
  • Chen Huang,

    Roles Data curation

    Affiliations Tea Research Institute, Chinese Academy of Agricultural Sciences, Hangzhou, Zhejiang, China, Key Laboratory of Tea Biology and Resources Utilization, Ministry of Agriculture, Hangzhou, Zhejiang, China

    ⨯
  • Xiangjie Hou,

    Roles Data curation

    Affiliations Tea Research Institute, Chinese Academy of Agricultural Sciences, Hangzhou, Zhejiang, China, Key Laboratory of Tea Biology and Resources Utilization, Ministry of Agriculture, Hangzhou, Zhejiang, China

    ⨯
  • Xiaoling Sun ,

    Roles Funding acquisition, Methodology, Project administration, Resources, Visualization, Writing – review & editing

    xlsun@mail.tricaas.com (XS); 1647785720@qq.com (BX)

    Affiliations Tea Research Institute, Chinese Academy of Agricultural Sciences, Hangzhou, Zhejiang, China, Key Laboratory of Tea Biology and Resources Utilization, Ministry of Agriculture, Hangzhou, Zhejiang, China

    ⨯
  • Bin Xiao

    Roles Supervision, Validation

    xlsun@mail.tricaas.com (XS); 1647785720@qq.com (BX)

    Affiliation College of Horticulture, Northwest A&F University, Yangling, Shaanxi, China

    ⨯

Abstract

Empoasca onukii Matsuda is one of the most devastating pests of the tea plant (Camellia sinensis). Still, the presumed expression stability of its reference genes (RGs) has not been analyzed. RGs are essential for accurate and reliable gene expression analysis, so this absence has hampered the study of the insect’s molecular biology. To find candidate RGs for normalizing gene expression data, we cloned ten common housekeeping genes from E. onukii. Using the ΔCt method, geNorm, NormFinder and BestKeeper, we screened the RGs that were appropriate for quantifying the mRNA transcription of cellular responses under five experimental conditions. We identified the combinations of α-TUB and G6PDH, α-TUB and UBC, two RGs (α-TUB and β-TUB1) or three RGs (α-TUB, RPL13 and GAPDH), AK and UBC, or RPL13 and α-TUB as the best for analyzing gene expression in E. onukii adults of both sexes in different tissues, nymphs at different developmental stages, nymphs exposed to different temperatures or nymphs exposed to photoperiod stress. Finally, the E. onukii cysteine proteinase (Eocyp) was chosen as the target gene to validate the rationality of the proposed RGs. In conclusion, our study suggests a series of RGs with which to study the gene expression profiles of E. onukii that have been manipulated (biotically or abiotically) using reverse transcription quantitative polymerase chain reaction. The results offer a solid foundation for further studies of the molecular biology of E. onukii.

Citation: Yu Y, Zhang J, Huang C, Hou X, Sun X, Xiao B (2018) Reference genes selection for quantitative gene expression studies in tea green leafhoppers, Empoasca onukii Matsuda. PLoS ONE 13(10): e0205182. https://doi.org/10.1371/journal.pone.0205182

Editor: Yulin Gao, Chinese Academy of Agricultural Sciences Institute of Plant Protection, CHINA

Received: June 8, 2018; Accepted: September 20, 2018; Published: October 8, 2018

Copyright: © 2018 Yu et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the paper and its Supporting Information files.

Funding: The study was sponsored by the Special Fund for Agro-scientific Research in the Public Interest (201403030), National Natural Science Foundation of China (31772180; 31471784).

Competing interests: The authors have declared that no competing interests exist.

Introduction

Reverse transcription quantitative polymerase chain reaction (RT-qPCR, hereafter qPCR) is a popular and indispensable technique for quantifying the mRNA transcription of cellular responses triggered by biotic or abiotic manipulations [1]. qPCR offers high-throughput screening, and is known to be fast, sensitive and accurate [2–4]. Every step of qPCR sample preparation and processing—determining the intrinsic variability of RNA, removing impurities during RNA extraction, carrying out reverse transcription and measuring PCR efficiencies—needs to be accurately normalized [5–7]. For now, stably expressed reference genes (RGs) are the best internal controls when results are quantified using the 2-ΔΔCt method or its modified versions [8, 9]. The measurement of internal controls along with target genes helps to compensate for the inevitable experimental variations, such as disparities in the amount of starting material and/or sample loading [10, 11]. Therefore, identifying appropriate RGs for a normalization scalar is an essential prerequisite for developing a qPCR assay. Furthermore, at least two RGs (preferably more) should be employed simultaneously in the normalization process [12–15].

Ideally, RGs should display constitutive and stable expression characteristics across cell lines, tissue types, developmental stages or experimental treatments, and should also be expressed at levels similar to those of target genes [5, 6]. Appropriate RGs are referred to as housekeeping genes. Although glyceraldehyde-3-phosphate (GAPDH), 18S ribosomal RNA (18S rRNA) and β-actin (ACTB) have been widely used in gene expression assays of invertebrates [16–19], increasing evidence has demonstrated that these genes have been used without proper validation [12, 20, 21]. Recent experiments have determined the most appropriate RGs for manipulations (biotic and abiotic) of the following species: Drosophila melanogaster (Meigen), Plutella xylostella (Linnaeus), Bemisia tabaci Mediterranean, Spodoptera litura (Fabricius), Sesamia inferens (Walker), S. exigua (Hübner), Bactrocera minax (Enderlein), Helicoverpa armigera Hübner, Myzus persicae (Sulzer), Nilaparvata lugens (Stål) and Galeruca daurica (Joannis) [22–32]. Results from these studies suggest that although RG expression is sometimes constant, it may vary considerably according to experimental treatment or to species. For example, arginine kinase (AK) was ranked as the most stable gene when N. lugens was exposed to temperature-induced stress or examined at different developmental stages, but AK was ranked as the most variable gene in the different tissues of the same species [33]. Moreover, ACTB in S. litura was ranked as the second most stable expression gene when insects were treated with insecticide but it was ranked as the most variable one when S. litura were fed on different foods or when we looked at different tissues [23]. By identifying the proper RGs for a given species under specific experimental conditions, we can avoid missing or overemphasizing potential biological changes in the expression of target genes.

The tea green leafhopper, Empoasca onukii Matsuda, one of the most devastating pests of the tea plant (Camellia sinensis (L.) O. Kuntze), usually produces ten generations per year in China [34, 35]. Both nymphs and adults of E. onukii attack the buds, tender leaves and stems of tea plants, causing plant parts to curl and parch [36]. Outbreaks of this hard-to-control insect can cause economic losses of up to 33% due to diminished tea yields [37]. A few molecular studies focusing on species assignment or transcriptome have investigated E. onukii [36, 38, 39] but little attention has been paid to gene expression analysis or even to the presumed expression stability of RGs. In order to obtain candidate RGs that were appropriate for quantifying the mRNA transcription of cellular responses under five experimental conditions, we cloned ten common RGs with different functions from E. onukii. Next, using qPCR, we measured ten mRNA transcriptional levels (ribosomal protein L13 (RPL13), alpha tubulin (α-TUB), ubiquitin conjugating enzyme (UBC), glutathione-S-transferase (GST), glyceraldehyde-3-phosphate dehydrogenase (GAPDH), TATA-box binding protein (TBP), glucose-6-phosphate dehydrogenase (G6PDH), AK and two β-tubulins (β-TUB1 and β-TUB2) in the whole bodies, in tissues from different body parts in male or female adults, in nymphs at different life stages and, finally, in nymphs treated with both temperature-induced stress and photoperiod-induced stress. The results were evaluated by using BestKeeper, geNorm, NormFinder and the ΔCt method to identify the most stably expressed RGs [12, 40–42]; an online tool, RefFinder, was then used to integrate the results to find the most stable. Finally, to demonstrate the importance of stable scalar in the normalization process of E. onukii gene expression, the E. onukii cysteine proteinase, Eocyp, which is expressed in all tissues and at all stages [43], was chosen. Our results identify a series of RGs that could be used with qPCR to study the gene expression profiles of E. onukii treated with induced stresses (biotic or abiotic); these will provide a solid foundation for further studies of the molecular biology of E. onukii.

Methods and materials

Experimental insects

Mixed-age E. onuki adults were collected from a tea plantation at the Tea Research Institute of the Chinese Academy of Agricultural Sciences (N 30°10', E 120°5'), Hangzhou, China. Adults were transported to freshly potted tea shoots in enclosed net cages (75 × 75 × 75 cm) and kept in a controlled climate room that was programmed at 26±2°C, 70±5% r.h., under a photoperiod of 14:10 h (L:D). Female adults laid eggs on the tender stems. After one generation, nymphs at different developmental stages or male and female adults were used for different treatments. All samples were frozen instantaneously with liquid nitrogen and stored in a -80°C refrigerator until use. Three biological replicates of all the treatments were prepared. The treatments are briefly summarized below (Table 1).

thumbnail
Download:
Table 1. Treatments and results.

https://doi.org/10.1371/journal.pone.0205182.t001

Developmental stages of nymphs.

Five treatment groups, composed of nymphs from the first through the fifth instars, were established. As some groups had few members, large numbers of nymphs at the same developmental stage (70 first-instar, 50 second-instar, 30 third-instar, 20 fourth-instar and 20 fifth-instar) were pooled separately for RNA extraction. Nymphs were separated by morphological characteristics under the microscope (Olympus SZ61, Beijing, China) and collected on the first day of molt.

Sexes.

Ten two-day-old virgin adult males and females were collected and used for separate analyses.

Tissues.

Two-day-old virgin male or female adults were dissected (head, thorax, and abdomen) by micro-forceps in liquid nitrogen under the microscope. Tissue from 20 male or female adults was collected and pooled as one sample.

Temperatures.

Ten fifth-instar nymphs (newly molted) were pooled in a glass tube (diameter 1.5cm, length 8.2cm) with one fresh tea shoot per sample. Tubes were then exposed for 1h to three temperature gradients in a metal bath: low (4°C), moderate (26°C) and high (36°C). The nymphs were allowed to recover at 26°C for another hour. Three sample pools (ten nymphs each) of each treatment were collected separately for RNA extraction. There was no mortality in response to the treatment.

Photoperiods.

The stability of candidate RGs was tested in fourth-instar nymphs subjected to three different photoperiod treatments in illuminated incubators (SenXinRGQ-360N, Shanghai, China) that were programmed at 26±2°C, 70±5% r.h.: 0:24 h (L:D), 14:10 h (L:D) and 24:0 h (L:D). Two days later, three sample pools (ten nymphs each) of each treatment were collected separately for RNA extraction. There was no mortality in response to the photoperiod treatment.

Total RNA isolation and cDNA synthesis

Total RNA was isolated by Promega SV Total RNA Isolation System according to the manufacturer’s protocol. The quantity and quality of extracted RNA were confirmed with NANODROP 2000c (Thermo Scientific, Wilmington, DE, USA). Ratios of A260/280 ranged from 2.0 to 2.2, suggesting a high level of purity among all RNA samples. One μg of RNA was used to synthesize the first-strand complementary DNA fragment using PrimeScript RT Master Mix (perfect real time) (TaKaRa, Tokyo, Japan), according to the manufacturer’s protocol. The standard curves were created with a five-fold dilution series of cDNA as a template for each treatment using a liner regression model. The cDNA of all samples was stored at -20°C.

Cloning and sequence identification

Ten housekeeping genes (α-TUB, AK, GAPDH, β-TUB1, RPL13, GST, β-TUB2, TBP, G6PDH and UBC) were selected as candidate RGs from the transcriptome database of E. onukii. These genes were cloned and sequenced, then confirmed by BLASTX. Primer premier 5 was used for primer design to clone these genes. Full-length cDNA was amplified by PCR using primers shown in Table 2. Each reaction included 50 ng cDNA, 1 μl of each primer (10 μM) and 2X PrimeSTAR Max Premix (TaKaRa, Tokyo, Japan), and the volume was adjusted with nuclease-free water to 20 μl. The PCR program contained a preliminary step at 95°C for 5min, 40 cycles of denaturation at 95°C for 10 s, an annealing temperature for 6s and 58°C for 1–3 min (depending on the length of mRNA). PCR products were examined by gel electrophoresis, purified using a TaKaRa MiniBEST Agarose Gel DNA Extraction Kit Ver 4.0 (TaKaRa, Tokyo, Japan), connected to the pMD 19-T Vector Cloning Kit (TaKaRa, Tokyo, Japan), following the manufacturer’s protocol, and cloned in Escherichia coli; three of the bacteria solutions were then sent to Genscript (Nanjing, China) for sequencing. The obtained sequences were compared to the transcriptomic database to confirm the sequence prediction, and an online tool (https://www.genscript.com/tools/real-time-pcr-tagman-primer-design-tool) was used for qPCR primer design (Table 3).

thumbnail
Download:
Table 2. Sequence information of the Candidate Reference Genes.

https://doi.org/10.1371/journal.pone.0205182.t002

thumbnail
Download:
Table 3. Primers of Candidate Reference Genes in E.onukii for qPCR.

https://doi.org/10.1371/journal.pone.0205182.t003

qPCR analysis

qPCR reactions were performed using a fluorescent quantitative system (LightCycler 480 Ⅱ) based on Synergy Brands. The 10μl reaction system contained 5 μl FastStart Essential DNA Green Master, 0.5 μl of forward and reverse primers (10 μM) and 25 ng of first-strand complementary DNA. The PCR program for all the genes included an initial denaturation step for 10 min at 95°C, followed by 45 cycles of 15s at 95°C, 15s at 60°C and 12s at 72°C. Finally, a melting curve analysis from 60°C to 95°C was performed to confirm the specificity of PCR products. Expression levels of these genes were determined by the number of cycles needed for the amplification to reach a fixed threshold in the exponential phase of the PCR reaction. Triplicates were carried out for each sample.

Validation of selected reference genes

To demonstrate the importance of proper RGs in the analysis of gene expression profiles, Eocyp (Genbank accession number: MH036890) was selected as the target gene. Three different normalization factors (NFs) were calculated based on the geometric mean of the genes with the lowest and the highest geometric mean values as determined by RefFinder, and a single RG with the lowest or the highest geometric mean value. Raw Ct values were transformed to relative quantities using the ΔΔCt method.

Data analysis

The stability of the candidate RGs was evaluated by BestKeeper, geNorm, NormFinder, the ΔCt method and the online tool RefFinder (http://150.216.56.64/referencegene.php?type=reference). The ΔCt method, geNorm and NormFinder rely on transformed Ct values of (1+E) –ΔCt, while original Ct values were used in BestKeeper and RefFinder. All these methods can recommend the most stable RGs; geNorm can also compare the pair-wise variation (V) of one gene with others. The value of Vn/n+1 indicates the pair-wise variation between two sequential NF and the optimal number of RGs required for accurate normalization; these two variables show the optimal number of RGs. Comparisons among more than two samples were analyzed using one-way ANOVA (Tukey’s test); those between two samples were analyzed using Student’s t-test.

Results

Identification, amplification and PCR efficiency for qPCR of E. onukii reference genes

Screening by PCR using primers (Table 2) yielded a single amplicon of the expected size for each RG. Each amplicon was cloned, sequenced and annotated in the GenBank database as follows: α-TUB, AK, GAPDH, β-TUB1, RPL13, GST, β-TUB2, G6PDH and UBC. For each pair of primers, the single normalized melting peak suggested that each pair of the primers amplified a unique product. The amplification efficiency (E) values for the 10 RGs ranged from 95.6 to 109.5%, with R2 values from 0.9927 to 0.9997 (Table 3), making all assays suitable for quantitative analysis.

Expression profiles of Candidate Reference Genes

As shown in Fig 1, raw Ct values of all candidate RGs ranged from 14.78 (AK) to 30.75 (TBP). α-TUB (17.31), AK (17.52), GAPDH (17.92) and β-TUB1 (18.08) were the most abundant transcripts, reaching the threshold fluorescence peak after 18 cycles. UBC (24.12) and TBP (25.95) were expressed at the lowest levels.

thumbnail
Download:
Fig 1. Expression profiles of Candidate Reference Genes in E. onukii.

The expression level of RGs in all samples is documented in terms of the cycle threshold number (Ct value). The data are expressed as box-whisker plots; the short bar in the box refers to Ct mean value; the box represents the 25th–75th percentiles; the median is indicated by a bar across the box; the whiskers on each box represent the distribution of the Ct values; and the dark spots refer to extreme outliers.

https://doi.org/10.1371/journal.pone.0205182.g001

Developmental stages of Nymphs

The gene expression stability of ten candidate RGs from nymphs at different developmental stages was analyzed using geNorm, the ΔCt method, BestKeeper and NormFinder. Results showed that the gene stability ranking as analyzed by geNorm differed from the ranking as analyzed by the other three methods (Table 4). For example, approaches using the ΔCt method, BestKeeper or NormFinder across the developmental stages of all nymphs identified α-TUB and UBC as the most stable RGs, whereas the geNorm approach identified the most stable genes across all developmental stages of all nymphs as RPL13 and G6PDH. However, all four methods all identified GST as the most variable. According to results using RefFinder, the ranking from the most to the least stable RG was as follows: α-TUB >UBC >G6PDH >RPL13 >β-TUB2 >GAPDH >AK >TBP >β-TUB1 >GST (Fig 2A). With GeNorm (Fig 3), all pairwise variation (Vn/n+1) was below 0.15 (the recommended cut-off), indicating that the inclusion of an additional RG was unnecessary (Fig 3). Based on the ranking of the RGs by RefFinder, α-TUB and UBC were identified as the best combination for the developmental stages of nymph E. onukii.

thumbnail
Download:
Fig 2. Expression stability of Candidate Reference Genes in E. onukii.

The stability of RG expression was measured by RefFinder. A lower geometric mean value represents more stable expression.

https://doi.org/10.1371/journal.pone.0205182.g002

thumbnail
Download:
Fig 3. Optimal number of reference genes for the normalization of E. onukii under different experimental manipulations.

The pairwise variation (Vn/n + 1) was analyzed by geNorm software to determine the optimal number of RGs included in the qPCR analysis. Values less than 0.15 indicate that another RG will not significantly improve normalization.

https://doi.org/10.1371/journal.pone.0205182.g003

thumbnail
Download:
Table 4. Ranking of reference genes expression under different experimental treatments.

https://doi.org/10.1371/journal.pone.0205182.t004

Sexes

BestKeeper, geNorm, ΔCt method and NormFinder identified α-TUB and G6PDH as the most stable RGs, and TBP as the least stable RG (Table 4). According to results from RefFinder, the ranking from the most to the least stable was as follows: α-TUB >G6PDH >β-TUB1 >β-TUB2 >AK >UBC >GST > GAPDH >RPL13 >TBP (Fig 2B). Based on results of geNorm, two RGs were suggested. According to RefFinder, α-TUB and G6PDH were chosen as the best combination for normalizing the expression of E. onukii adults in different sexes.

Tissues

Analyses of the ΔCt method, NormFinder and geNorm divided the ten RGs into two groups, each with different tissues of female E. onukii: the group of more stably expressed genes contained α-TUB, RPL13 and GAPDH; the group of less stable expressed genes contained AK, UBC and TBP (Table 4). The BestKeeper analysis revealed that UBC, α-TUB and GAPDH are the most stably expressed genes. According to the results from RefFinder, the stability ranking from the most to the least was as follows: α-TUB >RPL13 >GAPDH >GST >UBC >G6PDH >β-TUB1 >β-TUB2 >TBP >AK (Fig 2C). Meanwhile, the most stable RGs in different tissues of male adults were as follows: BestKeeper and the ΔCt method identified α-TUB, NormFinder identified β-TUB1, and geNorm identified TBP. However, all four methods identified AK as the most unstable gene (Table 4). Using geNorm and RefFinder in different tissues of female and male adults of E. onukii, normalization required two (α-TUB and β-TUB1) and three (α-TUB, RPL13 and GAPDH) RGs (Fig 3).

Temperatures

The ΔCt method and geNorm ranked AK as the most stably expressed genes in nymphs exposed to different temperatures, whereas BestKeeper and NormFinder ranked UBC as the most stable RG. All analysis programs, except BestKeeper, ranked TBP as the most variable gene. RefFinder ranked the genes from the most to the least stable as follows: AK >UBC >GST >α-TUB >G6PDH >GAPDH >RPL13 >β-TUB2 >TBP >β-TUB1 (Fig 2E). Furthermore, geNorm and RefFinder analysis revealed that all the V values were below 0.15, which means AK and UBC are the best RG combination for gene expression analysis when E. onukii nymphs were exposed to different temperatures (Fig 3).

Photoperiods

The ΔCt method, NormFinder and geNorm ranked RPL13, α-TUB, UBC and β-TUB1 as the four most stably expressed genes, and ranked β-TUB2, GST and TBP as the three most variably expressed genes, when nymphs were exposed to different photoperiods. BestKeeper also ranked RPL13, α-TUB and UBC as more stable genes than others, except TBP. RefFinder ranked the genes from the most to the least stable: RPL13 >α-TUB >UBC >β-TUB1 >TBP >G6PDH >AK >GAPDH >GST >β-TUB2 (Fig 2F). Analysis by geNorm also revealed that all V values were below 0.15 (Fig 3). Thus, the RGs recommended for the nymphs of E. onukii exposed to photoperiod stress are RPL13 and α-TUB.

Validation of proposed reference genes

Eocyp was chosen as the target gene to validate the rationality of the proposed RGs. The expression level of Eocyp in the third instar was significantly higher than that in the first instar when normalized with the combination of α-TUB and UBC (NF 1–2, F = 7.997, P = 0.004) or α-TUB, UBC and G6PDH (NF 1–3, F = 10.498, P = 0.001) as RGs, but no significant difference was found when normalized with only one RG (NF1, F = 6.537, P = 0.007 or NF10, F = 3.215, P = 0.061) or the combination of the two unstable RGs, β-TUB1 and GST, NF (9–10, F = 4.469, P = 0.025) (Fig 4A). Analogously, the expression level of Eocyp was also not the same when normalized with more than one RG [NF (1–2), F = 10.703, P = 0.010; NF (1–3), F = 5.656, P = 0.042; NF (1–4), F = 6.706, P = 0.030; NF (9–10), F = 9.755, P = 0.013] compared with compounds normalized with only one RG (NF1, F = 5.896, P = 0.038 or NF10, F = 6.843, P = 0.028) in different tissues of female adults. Moreover, when normalized with NF (1–2) (F = 7.669, P = 0.011), the expression level of Eocyp was significantly higher in the head than in the abdomen, but no significant difference was found when Eocyp was normalized with other RGs (Fig 4C). When the combination of α-TUB and G6PDH (NF 1–2, F = 0.063, P = 0.889) or the combination of α-TUB, G6PDH and β-TUB1 (NF 1–3, F = 0.058, P = 0.469) was used for normalization, similar expression levels of Eocyp were observed in female and male adults of E. onukii. However, when normalized with TBP (NF10, F = 4.600, P = 0.013) or the combination of PRL13 and TBP (NF 9–10, F = 1.730, P = 0.040), the expression level of Eocyp was higher in female adults than that in male adults (Fig 4B). Similar results were observed in different tissues of male adults as well, when normalized with the highest geometric mean value of AK (NF10, F = 86.973, P = 0.000) or the combination of AK and UBC (NF (9–10), F = 42.887, P = 0.000), the expression level of Eocyp in the head was higher than that in the thorax, whereas it remained the same in the head and thorax when assessed with the most appropriate RG of α-TUB (NF1, F = 6.696, P = 0.017) or the combination of α-TUB and β-TUB1 (NF1-2, F = 7.669, P = 0.011) (Fig 4D). Moreover, the expression level of Eocyp under the two abiotic stresses showed no significant difference when normalized with different individual RGs or combinations of RGs (Fig 4E and 4F).

thumbnail
Download:
Fig 4. Validation of the gene stability measure.

Expression profiles of Eocyp under different experimental conditions using different RGs. A. Nymphs at different developmental stages; B. Sex; C. Different tissues in female E. onukii; D. Different tissues in male E. onukii; E. Fifth-instar nymphs exposed to different temperatures; F. Fifth-instar nymphs exposed to different photoperiod. Data are means±SE. One-way ANOVA (Tukey’s test) was used to analyze significant difference among treatments (A, C~F); different letters in the same color columns show the statistical difference, P<0.05. Two samples were compared using Student’s t-test (B); *, P<0.05.

https://doi.org/10.1371/journal.pone.0205182.g004

Discussion

Accurate normalization is required to minimize errors in qPCR and is a prerequisite for obtaining reliable gene-expression results, especially when the differences are subtle. Many methods to assess the expression stability of RGs have been developed but until now none has been completely reliable. In addition, little attention has been paid to the presumed expression stability of RGs and to gene expression analysis in E. onukii. As a result, the molecular biology of this pest is not well understood. In the present study, we cloned ten common RGs with different functions from E. onukii and synthesized four commonly used methods to obtain the most stable RGs; these are needed for quantifying the mRNA transcription of cellular responses in response to five experimental manipulations. The methods—BestKeeper, ΔCt, NormFinder and geNorm—are based on different algorithms [24, 44]. BestKeeper analyzes the stability of the RG individually, whereas the ΔCt method, NormFinder and geNorm analyze pairwise variation between two RGs [45]. In our study, the results from BestKeeper varied somewhat from the results of the other three methods in response to the same treatment. For example, BestKeeper ranked UBC as the most stable RG, whereas the other methods identified it as the most unstable in both sexes of E. onukii (Table4). The negative correlation between UBC (r = 0.0073) and the other RGs (Table 4) indicates that the results of BestKeeper were imprecise. A similar phenomenon was found earlier by Pfaffl (2001) [46]. The ranking of RGs by geNorm also differed somewhat from the ranking of other methods when the developmental stages of nymphs were considered (Table 4). Previous results showed that geNorm was not able to evaluate the expression stability of RGs that were co-regulated or expressed similarly; thus, using geNorm presupposed the selection of a pair of RGs [42]. According to Fig 3, if G6PDH were excluded, the geNorm ranking of RGs of nymphs at different developmental stages would be changed and the stability of RPL13 would be correspondingly decreased. However, because the results of NormFinder did not show this trend (S1 Fig), we inferred that geNorm emphasizes the comparisons. Therefore, in order to obtain reliable evaluation results, our study used RefFinder.

In our present study, α-TUB exhibited the most stable expression status in response to most biotic manipulations, including the selection of the developmental stages of nymphs and of different tissues in adults of both sexes and in the whole bodies of both sexes (Fig 2). Furthermore, α-TUB also exhibited the most stability in different tissues and in the whole bodies of adults of each individual sex and in different tissue from a mixed-sex group of adults (S2–S5 Tables); G6PDH, on the other hand, exhibited the most stability across different stages of nymphs and in both sexes (S1 Table). α-TUB, which is a cytoskeletal protein and functions in many physiological processes [47], has been ranked as the most stable RG in expression studies under different developmental conditions with N. lugens, Liriomyza trifolii, B. tabaci, Tetranychus cinnabarinus, Monochamus alternatus and Sogatella furcifera [48–51]. In addition, α-TUB was ranked as the most stable RG in different tissues of C. suppressalis, Galeruca daurica, M. alternatus, Rhodnius prolixus and Phenacoccus solenopsis [17, 32, 52]. However, α-TUB was regarded as a variable RG in response to different developmental conditions of Apis mellifera and Delphacodes kuscheli [53, 54]. In conclusion, although α-TUB was expressed stably in most cases, it should be tested across more species and in response to different treatments before being employed.

Specific RGs may not be completely stable in a species exposed to different experimental conditions or to evaluation by different methods. Sometimes the instability may be due to a gene’s function. AK, arginine kinase, the only phosphagen kinase in insects [55], has been found to be the most stable RG in N. lugens and C. suppressalis when these were subjected to temperature stresses or insecticides, and also the most stable RG in different larval tissues of S. litura, and in the labial gland and fat body of Bombus terrestris [56, 57]. Our results showed that AK was the most stable RG in E. onukii nymphs exposed to different temperatures (Fig 2E) but it was the most unstable RG in tissues from the head and/or the thorax of both male and female adults (Fig 2B and 2C). Ribosomal protein 13L (RPL13) encodes a ribosomal protein that is a component of the 60S subunit and, in conjunction with rRNA, constitutes the ribosomal subunits involved in the cellular process of translation [47]. Our results showed that RPL13 was an optimized RG of E. onukii in different tissues of female adults and in nymphs that had been exposed to different photoperiods (Fig 2F). Previous results showed that the ribosomal protein gene was found to be the most stable RG in A. mellifera, Schistocerca gregaria, Tribolium castaneum, D. melanogaster, Bombyx mori, C. suppressalis and B. tabaci under certain experimental conditions [22, 29, 53, 57–60]. GST was regarded as the most unstable RG in E. onukii nymphs at different developmental stages or in nymphs exposed to different photoperiods; however, RefFinder ranked GST as the most stable RG when all experimental conditions were considered (S6 Table). In conclusion, the normalization scalars among the candidates correspond to experimental conditions and the best way to select a suitable RG for gene expression analysis is to evaluate it under specific experimental conditions.

Recently, multiple RGs have been used to normalize the expression of target genes more accurately. When multiple RGs were used in a given experiment, the probability of biased normalization was reduced [19, 61–64]. Traditionally, an optimal number of RGs with the least pairwise variation (V) was selected by GeNorm. GeNorm determines the pairwise variations (Vn/n+1) in NFs (the geometric mean of multiple RGs) using n or n+1 RGs with a threshold value< 0.15 but the threshold value is not totally absolute [12]. In the present study, our results not only confirmed that the most appropriate RGs differed across experimental manipulations but they also proved that the use of multiple RGs in qPCR increased the accuracy and sensibility of gene expression analysis in E. onukii (Fig 4A–4D). Furthermore, the results demonstrated that when Eocyp expression data were normalized with the combination of α-TUB, RPL13 and GAPDH, subtle differences among different tissues of female adult E. onukii were detected but when data were normalized with one RG or with other combinations of RGs, no differences were observed.

In conclusion, using qPCR, we screened a series of RGs to study the gene expression profiles of E. onukii in response to multiple experimental manipulations. This study provide a solid foundation for further studies of the molecular biology of E. onukii.

Supporting information

S1 Table. Expression stability of Candidate Reference Genes across different developmental stages of Nymphs and between sexes.

https://doi.org/10.1371/journal.pone.0205182.s001

(DOCX)

S2 Table. Expression stability of Candidate Reference Genes in different tissues and whole bodies of E. onukii Adult Females.

https://doi.org/10.1371/journal.pone.0205182.s002

(DOCX)

S3 Table. Expression stability of the candidate RGs in different tissues and whole bodies of E. onukii adult males.

https://doi.org/10.1371/journal.pone.0205182.s003

(DOCX)

S4 Table. Expression stability of the Candidate Reference Genes in different tissues in E. onukii male and female adults.

https://doi.org/10.1371/journal.pone.0205182.s004

(DOCX)

S5 Table. Expression stability of the Candidate Reference Genes in different tissues and whole body in E. onukii male and female adults.

https://doi.org/10.1371/journal.pone.0205182.s005

(DOCX)

S6 Table. Expression stability of Candidate Reference Genes under all conditions.

https://doi.org/10.1371/journal.pone.0205182.s006

(DOCX)

S1 Fig. Robustness of geNorm and NormFinder.

To assess the robustness of the rankings among the candidate genes, we compared the results obtained using all genes (solid lines), the results obtained when excluding G6PDH (dashed lines) or α-TUB (dot lines) in response to the developmental stages of nymph E. onukii. The results described by geNorm use black curves; red curves represent the values described by NormFinder. The top ranked genes are those with the smallest values for each method.

https://doi.org/10.1371/journal.pone.0205182.s007

(DOCX)

Acknowledgments

We gratefully acknowledge Emily Wheeler for editorial assistance.

References

  1. 1. Ginzinger DG. Gene quantification using real-time quantitative PCR: An emerging technology hits the mainstream. Experimental Hematology. 2002;30: 503–512. https://doi.org/10.1016/S0301-472x(02)00806-8 pmid:12063017.
  2. 2. Gibson UE, Heid CA, Williams PM. A novel method for real time quantitative RT-PCR. Genome Research. 1996;6: 995–1001. pmid:8908519.
  3. 3. Williams PM, Livak KJ, Stevens J, Heid CA. Real time quantitative PCR. Genome Research. 1996;6: 986–994. pmid:8908518
  4. 4. Schefe JH, Lehmann KE, Buschmann IR, Unger T, Funkekaiser H. Quantitative real-time RT-PCR data analysis: current concepts and the novel "gene expression's CT difference" formula. Journal of Molecular Medicine. 2006;84: 901–910. pmid:16972087
  5. 5. Nijhof AM, Balk JA, Postigo M, Jongejan F. Selection of reference genes for quantitative RT-PCR studies in Rhipicephalus (Boophilus) microplus and Rhipicephalus appendiculatus ticks and determination of the expression profile of Bm86. BMC Molecular Biology. 2009;10: 1–14. PMID: 19126214.
  6. 6. Chapman JR, Waldenstrom J. With Reference to Reference Genes: A systematic review of endogenous controls in gene expression studies. PLoS One. 2015;10: e0141853. pmid:26555275.
  7. 7. Osman F, Hodzic E, Kwon SJ, Wang J, Vidalakis G. Development and validation of a multiplex reverse transcription quantitative PCR (RT-qPCR) assay for the rapid detection of Citrus tristeza virus, Citrus psorosis virus, and Citrus leaf blotch virus. Journal of Virological Methods. 2015;220: 64–75. pmid:25907469.
  8. 8. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2−ΔΔCtmethod. Methods. 2001;25: 402–408. pmid:11846609.
  9. 9. Schmittgen TD, Livak KJ. Analyzing real-time PCR data by the comparative CT method. Nature Protocols. 2008;3: 1101–1108. https://doi.org/10.1038/nprot.2008.73. pmid:18546601
  10. 10. VanGuilder HD, Vrana KE, Freeman WM. Twenty-five years of quantitative PCR for gene expression analysis. BioTechniques. 2008;44: 619–626. pmid:18474036.
  11. 11. Nestorov J, Matić G, Elaković I, Tanić N. Gene expression studies: how to obtain accurate and reliable data by quantitative real-time RT PCR. Journal of Medical Biochemistry. 2013;32: 325–338. https://doi.org/10.2478/jomb-2014-0001.
  12. 12. Vandesompele J, De Preter K, Pattyn F, Poppe B, Van Roy N, De Paepe A, et al. Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biology. 2002;3. pmid:12184808.
  13. 13. Sinha DK, Smith CM. Selection of Reference genes for expression analysis in Diuraphisnoxia (Hemiptera: Aphididae) fed on resistant and susceptible wheat plants. Scientific Reports. 2014;4: 5059. pmid:24862828.
  14. 14. Chandna R, Augustine R, Bisht NC. Evaluation of candidate reference genes for gene expression normalization in Brassica juncea using real time quantitative RT-PCR. PloS One. 2012;7: e36918. pmid:22606308.
  15. 15. Xiao X, Ma J, Wang J, Wu X, Li P, Yao Y. Validation of suitable reference genes for gene expression analysis in the halophyte Salicornia europaea by real-time quantitative PCR. Front in Plant Science. 2014;5: 788. pmid:25653658.
  16. 16. de Jonge HJ, Fehrmann RS, de Bont ES, Hofstra RM, Gerbens F, Kamps WA, et al. Evidence based selection of housekeeping genes. PLoS One. 2007;2: e898. pmid:17878933.
  17. 17. Paim RM, Pereira MH, Di Ponzio R, Rodrigues JO, Guarneri AA, Gontijo NF, et al. Validation of reference genes for expression analysis in the salivary gland and the intestine of Rhodnius prolixus (Hemiptera, Reduviidae) under different experimental conditions by quantitative real-time PCR. BMC Research Notes. 2012;5: 128. pmid:22395020.
  18. 18. Bagnall NH, Kotze AC. Evaluation of reference genes for real-time PCR quantification of gene expression in the Australian sheep blowfly, Lucilia cuprina. Medical and Veterinary Entomology.2010;24: 176–181. pmid:20604863.
  19. 19. Coleman AD, Wouters RH, Mugford ST, Hogenhout SA. Persistence and transgenerational effect of plant-mediated RNAi in aphids. Journal of Experimental Botany. 2015;66: 541–548. pmid:25403918.
  20. 20. Huang X, Gao Y, Jiang B, Zhou Z, Zhan A. Reference gene selection for quantitative gene expression studies during biological invasions: A test on multiple genes and tissues in a model ascidian Ciona savignyi. Gene. 2016;576: 79–87. pmid:26428313.
  21. 21. Shang F, Wei DD, Jiang XZ, Wei D, Shen GM, Feng YC, et al. Reference gene validation for quantitative PCR under various biotic and abiotic stress conditions in Toxoptera citricida (Hemiptera, Aphidiae). Journal of Economic Entomology. 2015;108: 2040–2047. pmid:26470351.
  22. 22. Ponton F, Chapuis MP, Pernice M, Sword GA, Simpson SJ. Evaluation of potential reference genes for reverse transcription-qPCR studies of physiological responses in Drosophila melanogaster. Journal of Insect Physiology. 2011;57: 840–850. pmid:21435341.
  23. 23. Lu YH, Yuan M, Gao XW, Kang TH, Zhan S, Wan H, et al. Identification and validation of reference genes for gene expression analysis using quantitative PCR in Spodoptera litura (Lepidoptera: Noctuidae). PloS One. 2013;8: e68059. pmid:23874494.
  24. 24. Fu W, Xie W, Zhang Z, Wang S, Wu Q, Liu Y, et al. Exploring valid reference genes for quantitative real-time PCR analysis in Plutella xylostella (Lepidoptera: Plutellidae). International Journal of Biological Science. 2013;9: 792–802. pmid:23983612.
  25. 25. Xun Z, Miao Y, Muhammad S, Zhang Y, Wang S, Xin W, et al. Selection and evaluation of reference genes for expression analysis using qRT-PCR in the beet armyworm Spodoptera exigua(Hübner) (Lepidoptera: Noctuidae). PloS One. 2014;9: e84730. pmid:24454743.
  26. 26. Lü ZC, Wang LH, Dai RL, Zhang GF, Guo JY, Wan FH. Evaluation of endogenous reference genes of Bactrocera (Tetradacus) minax by gene expression profiling under various experimental conditions. Florida Entomologist. 2014;97: 597–604. https://doi.org/10.1653/024.097.0235.
  27. 27. Sun M, Lu MX, Tang XT, Du YZ. Exploring valid reference genes for quantitative real-time PCR analysis in Sesamia inferens (Lepidoptera: Noctuidae). PLoS One. 2015;10: e0115979. pmid:25585250.
  28. 28. Zhang SD, An SH, Li Z, Wu FM, Yang QP, Liu YC, et al. Identification and validation of reference genes for normalization of gene expression analysis using qRT-PCR in Helicoverpa armigera (Lepidoptera: Noctuidae). Gene. 2015;555: 393–402. pmid:25447918.
  29. 29. Dai TM, Lu ZC, Liu WX, Wan FH. Selection and validation of reference genes for qRT-PCR analysis during biological invasions: the thermal adaptability of Bemisia tabaciMED. PLoS One. 2017;12: e0173821. pmid:28323834.
  30. 30. Zhi-Wei Kang, Fang-Hua Hong-Gang, Tian Meng, et al. Evaluation of the reference genes for expression analysis using quantitative real-time polymerase chain reaction in the green peach aphid, Myzus persicae. Insect Science. 2016;24: 222–234. pmid:26749166
  31. 31. Wan PJ, Tang YH, Yuan SY, He JC, Wang WX, Lai FX, et al. Reference genes for quantitative real-time PCR analysis in symbiont Entomomyces delphacidicola of Nilaparvata lugens (Stal). Scientific Reports. 2017;7: 42206. pmid:28198810.
  32. 32. Tan Y, Zhou XR, Pang BP. Reference gene selection and evaluation for expression analysis using qRT-PCR in Galeruca daurica (Joannis). Bulletin Entomological Research. 2017;107: 359–368. pmid:27819206.
  33. 33. Yuan M, Lu YH, Zhu X, Wan H, Shakeel M, Zhan S, et al. Selection and evaluation of potential reference genes for gene expression analysis in the brown planthopper, Nilaparvata lugens (Hemiptera: Delphacidae) using reverse-transcription quantitative PCR. PloS One. 2014;9: e86503. pmid:24466124.
  34. 34. Jin S, Chen ZM, Backus EA, Sun XL, Xiao B. Characterization of EPG waveforms for the tea green leafhopper, Empoasca vitis Gothe (Hemiptera: Cicadellidae), on tea plants and their correlation with stylet activities. Journal of Insect Physiology. 2012;58: 1235–1244. pmid:22750027.
  35. 35. Wei Q, Yu HY, Niu CD, Yao R, Wu SF, Chen Z, et al. Comparison of insecticide susceptibilities of Empoasca vitis (Hemiptera: Cicadellidae) from three main tea-growing regions in China. Journal of Economic Entomology. 2015;108: 1251. pmid:26470253.
  36. 36. Shao ES, Lin GF, Liu S, Ma XL, Chen MF, Lin L, et al. Identification of transcripts involved in digestion, detoxification and immune response from transcriptome of Empoasca vitis (Hemiptera: Cicadellidae) nymphs. Genomics. 2017;109: 58–66. pmid:27867104.
  37. 37. Xu JH, Wang NW, Zhang LL, Guan X. Study on the economic threshold of tea leafhopper (Empoasca vitis Gothe). Journal of Tea Science. 2005. https://doi.org/10.13305/j.cnki.jts.2005.02.009.
  38. 38. Fu JY, Han BY, Xiao Q. Mitochondrial COI and 16sRNA evidence for a single species hypothesis of E. vitis, J. formosana and E. onukii in East Asia. PLoS One. 2014;9: e115259. pmid:25506929.
  39. 39. Qin DZ, Zhang L, Xiao Q, Dietrich C, Matsumura M. Clarification of the identity of the tea green leafhopper based on morphological comparison between Chinese and Japanese specimens. PloS One. 2015;10: e0139202. pmid:26422616.
  40. 40. Silver N, Best S, Jiang J, Thein SL. Selection of housekeeping genes for gene expression studies in human reticulocytes using real-time PCR. BMC Molecular Biology. 2006;7: 33. pmid:17026756.
  41. 41. Pfaffl MW, Tichopad A, Prgomet C, Neuvians TP. Determination of stable housekeeping genes, differentially regulated target genes and sample integrity: BestKeeper–Excel-based tool using pair-wise correlations. Biotechnological Letters. 2004;26: 509–515.
  42. 42. Andersen CL, Jensen JL, Orntoft TF. Normalization of real-time quantitative reverse transcription-PCR data: a model-based variance estimation approach to identify genes suited for normalization, applied to bladder and colon cancer data sets. Cancer Research. 2004;64: 5245–5250. pmid:15289330.
  43. 43. YU Y, SUN X, Xiao B. Molecular cloning and expression of a cysteine gene in Empoasca onukii. Journal of Tea Science. 2018. https://doi.org/10.13305/j.cnki.jts.2018.03.006.
  44. 44. Yang CX, Pan HP, Liu Y, Zhou XG. Selection of reference genes for expression analysis using quantitative real-time PCR in the pea aphid, Acyrthosiphon pisum (Harris) (Hemiptera, Aphidiae). PloS One. 2014;9: e110454. pmid:25423476.
  45. 45. Teng XL, Zhang Z, He GL, Yang LW, Li F. Validation of reference genes for quantitative expression analysis by real-time RT-PCR in four lepidopteran insects. Journal of Insect Science. 2012;12: 1–17. pmid:22938136.
  46. 46. Pfaffl MW. A new mathematical model for relative quantification in real-time RT–PCR. Nucleic Acids Research. 2001;29: e45. pmid:11328886.
  47. 47. Campbell MG, Karbstein K. Protein-protein interactions within late pre-40S ribosomes. PloS One. 2011;6: e16194. pmid:21283762.
  48. 48. Sun W, Jin Y, He L, Lu WC, Li M. Suitable reference gene selection for different strains and developmental stages of the carmine spider mite, Tetranychus cinnabarinus, using quantitative real-time PCR. Journal of Insect Science. 2010;10: 208. pmid:21265619.
  49. 49. Feng B, Guo QS, Mao BP, Yong-Jun DU. Identification and selection of valid reference genes for assaying gene expression in the chemosensory tissues of Monochamus alternatus (Coleoptera: Cerambycidae) by RT-qPCR. Acta Ecologica Sinica. 2016;59: 427–437. https://doi.org/10.16380/j.kcxb.2016.04.008.
  50. 50. An XK, Hou ML, Liu YD. Reference gene selection and evaluation for gene expression studies using qRT-PCR in the white-backed planthopper, Sogatella furcifera (Hemiptera: Delphacidae). Journal of Economic Entomology. 2016;109: 879–886. pmid:26612891.
  51. 51. Chang YW, Chen JY, Lu MX, Gao Y, Tian ZH, Gong WR, et al. Selection and validation of reference genes for quantitative real-time PCR analysis under different experimental conditions in the leafminer Liriomyza trifolii (Diptera: Agromyzidae). PLoS One. 2017;12: e0181862. pmid:28746411.
  52. 52. Chen F, Yong-Yue LU, Entomology DO. Selection of reference genes in Phenacoccus solenopsis (Hemiptera:Pseudococcidae) under heat stress. Acta Ecologica Sinica. 2014;57: 1146–1154. https://doi.org/10.16380/j.kcxb.2014.10.008.
  53. 53. Lourenco AP, Mackert A, Cristino AD, Simoes ZLP. Validation of reference genes for gene expression studies in the honey bee, Apis mellifera, by quantitative real-time RT-PCR. Apidologie. 2008;39: 372–385. PMID: 27182699.
  54. 54. Maroniche GA, Sagadin M, Mongelli VC, Truol GA, del Vas M. Reference gene selection for gene expression studies using RT-qPCR in virus-infected planthoppers. Virology Journal. 2011;8: 308. pmid:21679431.
  55. 55. Tanaka K, Ichinari S, Iwanami K, Yoshimatsu S, Suzuki T. Arginine kinase from the beetle Cissites cephalotes (Olivier). Molecular cloning, phylogenetic analysis and enzymatic properties. Insect Biochemistry and Molecular Biology 2007;37: 338–345. pmid:17368197.
  56. 56. Hornakova D, Matouskova P, Kindl J, Valterova I, Pichova I. Selection of reference genes for real-time polymerase chain reaction analysis in tissues from Bombus terrestris and Bombus lucorum of different ages. Analytical Biochemistry. 2010;397: 118–120. pmid:19751695.
  57. 57. Xu J, Lu MX, Cui YD, Du YZ. Selection and evaluation of reference genes for expression analysis using qRT-PCR in Chilo suppressalis (Lepidoptera: Pyralidae). Journal of Economic Entomology. 2017;110: 683–691. pmid:28115499.
  58. 58. Van Hiel MB, Van Wielendaele P, Temmerman L, Van Soest S, Vuerinckx K, Huybrechts R, et al. Identification and validation of housekeeping genes in brains of the desert locust Schistocerca gregaria under different developmental conditions. BMC Molecular Biology. 2009;10: 56. pmid:19508726.
  59. 59. Peng R, Su BJ, Zhao GD, Ji X, Zhao SS, Zhang T, et al. Standardization of reference genes in silkworm, Bombyx mori. Advanced Materials Research. 2011;175–176: 67–71. https://doi.org/10.4028/www.scientific.net/AMR.175-176.67.
  60. 60. Liu JB, Ou J, Yao FJ, He L, Xi GS, Wei ZM. Identification of appropriate reference genes for gene expression studies by quantitative real-time PCR in Tribolium castaneum after exposure to phosphine. Journal of Agricultural Biotechnology. 2014;22: 257–264. https://doi.org/10.3969/j.issn.1674-7968.2014.02.016.
  61. 61. Tatsumi K, Ohashi K, Taminishi S, Okano T, Yoshioka A, Shima M. Reference gene selection for real-time RT-PCR in regenerating mouse livers. Biochemical and Biophysical Research Communications. 2008;374: 106–110. pmid:18602371.
  62. 62. Shakesby AJ, Wallace IS, Isaacs HV, Pritchard J, Roberts DM, Douglas AE. A water-specific aquaporin involved in aphid osmoregulation. Insect Biochemistry and Molecular Biology. 2009;39: 1–10. pmid:18983920.
  63. 63. Veazey KJ, Golding MC. Selection of stable reference genes for quantitative rt-PCR comparisons of mouse embryonic and extra-embryonic stem cells. PLoS One. 2011;6: e27592. pmid:22102912.
  64. 64. Pitino M, Coleman AD, Maffei ME, Ridout CJ, Hogenhout SA. Silencing of aphid genes by dsRNA feeding from plants. PloS One. 2011;6: e25709. pmid:21998682.