Introduction
Thrips are major pests of many agricultural crops throughout the world. They directly damage the host plants by sucking fluids from buds, leaves, flowers, fruits, and twigs resulting in distortions, stunted growth, feeding scars, and color mosaicism. Moreover, thrips are the sole vectors of plant Tospovirus. For example, onion thrips, Thrips tabaci Lindeman, can transmit Tospovirus as well as Iris yellow spot virus causing more than US$100 million losses every year (Prins & Goldbach, Reference Prins and Goldbach1998; Gent et al., Reference Gent, du Toit, Fichtner, Mohan, Pappu and Schwartz2006). Traditionally, identification of thrips, which are minute in size and have a high degree of similarity in appearance, is mainly based on adult characters. Moreover, some cryptic species, such as Frankliniella occidentalis (Pergande), Scirtothrips dorsalis Hood, and T. tabaci, which exhibit large differences in genetic compositions, habitat preference, host plant, and Tospovirus transmission efficiency, are virtually indistinguishable with morphology (Brunner et al., Reference Brunner, Chatzivassiliou, Katis and Frey2004; Toda & Murai, Reference Toda and Murai2007; Hoddle et al., Reference Hoddle, Heraty, Rugman-Jones, Mound and Stouthamer2008; Brunner & Frey, Reference Brunner and Frey2010; Rugman-Jones et al., Reference Rugman-Jones, Hoddle and Stouthamer2010; Jacobson et al., Reference Jacobson, Booth, Vargo and Kennedy2013).
PCR amplicons from thrips’ genomic DNA are commonly used for thrips identification and phylogenetic analysis including the nuclear ribosomal DNA and elongation factor (EF1-α) (Inoue & Sakurai, Reference Inoue and Sakurai2007; Hoddle et al., Reference Hoddle, Heraty, Rugman-Jones, Mound and Stouthamer2008; Buckman et al., Reference Buckman, Mound and Whiting2013) and mitochondrial DNA, e.g., COI and 16S rDNA, (Lin et al., Reference Lin, Wang and Yeh2003; Brunner et al., Reference Brunner, Chatzivassiliou, Katis and Frey2004; Asokan et al., Reference Asokan, Kumar, Kumar and Ranganath2007; Toda & Murai, Reference Toda and Murai2007; Hoddle et al., Reference Hoddle, Heraty, Rugman-Jones, Mound and Stouthamer2008). Most of the above mentioned studies have elucidated the phylogenetic relationships among species in a given genus and provided reliable tools for species identification. For example, COI and 28S rDNA sequence data showed that S. dorsalis consists of at least three distinct taxa (Hoddle et al., Reference Hoddle, Heraty, Rugman-Jones, Mound and Stouthamer2008). The internal transcribed spacer (ITS), the non-coding fragment of the nuclear ribosomal region, has been one of the most widely used markers in thrips species identification and population delineation (Liu, 2004; Rugman-Jones et al., Reference Rugman-Jones, Hoddle, Mound and Stouthamer2006; Farris et al., Reference Farris, Ruiz-Arce, Ciomperlik, Vasquez and DeLeón2010).
PCR-based methods, such as species-specific primer assay (Liu, Reference Liu2004; Asokan et al., Reference Asokan, Kumar, Kumar and Ranganath2007; Farris et al., Reference Farris, Ruiz-Arce, Ciomperlik, Vasquez and DeLeón2010; Kobayashi & Hasegawa, Reference Kobayashi and Hasegawa2012), restriction fragment length polymorphism (Lin et al., Reference Lin, Wang and Yeh2003; Rugman-Jones et al., Reference Rugman-Jones, Hoddle, Mound and Stouthamer2006), and real-time PCR (Walsh et al., Reference Walsh, Boonham, Barker and Collins2005; Huang et al., Reference Huang, Lee, Yeh, Shen, Mei and Chang2010) have been widely applied to thrips identification. However, these methods have been focused only on the identification of one or a few thrips species. It is essential to develop a more efficient method for simultaneous screening of mass samples. In the past decade, the microarray assay routinely used in pathogen investigation has been used rarely for insect pest identification (Chung et al., Reference Chung, Kang, Kim, Kim, Jung, Lee, Lee and Hwang2011; Yeh et al., Reference Yeh, Lee, Tseng, Chang and Wu2012; de Luca et al., Reference de Luca, Ribeiro, Sakai, Muraosa, Lyra, Gonoi, Mikami, Tominaga, Kamei, Zaninelli, Trabasso and Moretti2013; Lee et al., Reference Lee, Choi, Kang, Kim, Lee, Lee and Hwang2013).
Many studies have shown that sequence variation in the COI gene within thrips species is generally less than 2%, yet most of these studies did not have comparable data for ITS sequences (Brunner et al., Reference Brunner, Chatzivassiliou, Katis and Frey2004; Asokan et al., Reference Asokan, Kumar, Kumar and Ranganath2007; Rugman-Jones et al., Reference Rugman-Jones, Hoddle and Stouthamer2010; Kobayashi & Hasegawa, Reference Kobayashi and Hasegawa2012; Kadirvel et al., Reference Kadirvel, Srinivasan, Hsu, Su and de la Peña2013). Glover et al. (Reference Glover, Collins, Walsh and Boonham2010) have pointed out that as compared to an average sequence distance of 23.1% for the COI gene, an average interspecific distance of 54% in the hypervariable ITS region would have a significant advantage for species-level identification in thrips. In this study, therefore, species-specific primers were designed based on the established ITS1 sequences of 15 agriculturally important thrips, including T. tabaci, Thrips hawaiiensis (Morgan), Thrips palmi Karny, S. dorsalis, Frankliniella intonsa (Trybom), and F. occidentalis. The specificity and stability of these primers were examined on a total of 16 thrips species. Moreover, a DNA microarray based on these verified specific sequences provides an efficient and high-throughput method for thrips identification and monitoring. This microarray assay technique using ITS sequences could be widely applied to the rapid identification of large numbers of other agricultural pests as well.
Materials and methods
Thrips specimens were collected between 2004 and 2009 from localities across Taiwan and preserved in 95% ethanol. Fifteen thrips commonly found on agricultural crops, including virus transmission species such as T. tabaci, F. intonsa, and S. dorsalis, were used to develop the specific primers and probes. Additionally, eight thrips were employed for primer specificity examination. Pertinent information for these thrips species is given in table 1.
Table 1. Thrips species and their abbreviations in this study. Thrips used to design species-specific primers are in bold and others were included for specificity and stability examination. Arabic numerals are the representative electrophoresis lanes in Figure 2.
DNA extraction
Total DNA was extracted from individual thrips using the BuccalAmp™ DNA Extraction Kit (EPICENTRE Biotechnologies, Madison, USA) with instructions modified for thrips (Tseng et al., Reference Tseng, Chang, Tseng and Yeh2010). Individual specimens of the 23 thrips species, listed in table 1, were immersed in 50 μl DNA Extraction Solution 1.0. After shaking vigorously for 15 s, the sample was incubated at 65 °C for 15–20 min, followed by an additional 15 s of shaking. After removing the specimen, the reaction mixture was incubated at 98 °C for 2 min and then stored at −20 °C. The specimen was subsequently mounted on slide via Hoyer's medium for identification (Han, Reference Han1997; Mound & Kibby, Reference Mound and Kibby1998; Wang, Reference Wang2002, Reference Wang2007) and these voucher specimens are stored at the Laboratory of Molecular Systematics, Department of Entomology, National Chung Hsing University.
PCR and DNA sequencing
Primer pairs used for ITS1 fragment amplification and sequencing were 18Se (5′TCCCTGCCCTTTGTACACAC3′) and 5.8SThR (5′CACAAGCCRAGGGATCCAC3′), which were designed in this study based on conserved fragments of 18S rDNA and 5.8S rDNA (Tautz et al., Reference Tautz, Hancock, Webb, Tautz and Dover1988; Kjer et al., Reference Kjer, Baldridge and Fallon1994). PCR assay was carried out in a volume of 25 μl, with the following programming conditions: 95 °C for 2 min for the first denaturation, followed by 35 cycles of 94 °C for 40 s, 50 °C for 50 s and 72 °C for 1 min, with a final extension at 72 °C for 10 min. The amplified product was purified directly using a PCR purification kit (Quiagen, Hilden, German), or after resolving on agarose gel, excised and extracted with the Qiaquick gel extraction kit. The resulting DNA product was sequenced in both directions using BigDye Terminator V3.1 Cycle Sequencing Kit and an ABI 3730XL sequencer (Applied Biosystems, California, USA).
Sequence analysis
Forty-three ITS1 sequences of the 15 target thrips species were aligned with 28 sequences of 21 thrips retrieved from GenBank, including Echinothrips americanus Morgan, F. intonsa, F. occidentalis, Frankliniella schultzei (Trybom), Neohydatothrips geminus (Hood), Neohydatothrips burungae (Hood), Haplothrips chinensis Priesner, and 14 Scirtothrips species, using the program MAFFT (Katoh et al., Reference Katoh, Kuma, Toh and Miyata2005) or MUSCLE (Edgar, Reference Edgar2004) and manual editing. Pairwise distance was estimated using uncorrected proportional divergence with MEGA5 (Tamura et al., Reference Tamura, Peterson, Peterson, Stecher, Nei and Kumar2011).
Design of species-specific primers and multiplex PCR
Species-specific primers were developed from the variable regions, i.e., sequences that could not be adjusted to be conserved among thrips species, including those acquired in this study and those from GenBank. Two species-specific primer pairs with Tm around 60 °C and a size of 20 to 30 bp were designed for each of the target thrips (table 2), and their specificity and stability were examined on 16 thrips species of different genera (table 1). A multiplex PCR method was adopted using the species-specific primers combined with one universal primer pair. PCR conditions for testing the specificity and stability of these primer sets were the same as those employed in ITS1 amplification, except that the extension time (at 72 °C) was shortened from 50 to 30 s. Moreover, the universal primer pair 28Sg and 28Sh (Lin et al., Reference Lin, Wang and Yeh2003) were used in each multiplex reaction. Products were visualized on agarose gel.
Table 2. Sequences of species-specific primers and probes and their amplified fragments size for 15 thrips species. Abbreviations of thrips species are the same as in Table 1. Alternative specific probes for four thrips species are shown in footnote (see Materials and methods).
* Panel in fig. 2
Dsmi1F: 5′TCTGTGGTTCGAATAAGTCCC3′; Dsmi3R: 5′ATTTTTGTTTGCCCGACTCCC3′; Focc2F: 5′TCAGAGACGGTTCGATTCC3′; Sbif3F: 5′TGGGGCCTGAACTCGAATC3′; Sbif3R: 5′TTTCGGCGCGTTATAAACGC3′; Tpal3F: 5′ACGAACCGAAAGACGAGAAAC3′; Tpal4F: 5′TGCTTCCAAGTTCTTCGAAGG3′.
Probe design
Based on the verified specific primers in multiplex PCR, species-specific oligonucleotide probes with 5′ biotin labeling were synthesized. One control probe, i.e., Thrips-I1-1U, from the DR Thrips™ C8 Kit (DR. Chip Biotechnology Inc., Taiwan) was used to confirm normal hybridization, and two universal thrips probes, i.e., 18Se and 5.8SThR, served as positive controls. When the specific probes yielded a weak signal or showed cross hybridization, alternative specific probes, i.e., Dsmi1F, Dsmi3R, Focc2F, Sbif3F, Sbif3R, Tpal3F and Tpal4F, were used (table 2).
Microarray chip construction
The polymer substrate and colorimetric reagents in microarray test were provided by DR Chip DIY™ Kit (DR. Chip Biotechnology Inc., Taiwan). The probe solution (20 μM), prepared by mixing the 40 μM oligonucleotide probe with 2× probe solution, was spotted on the surface of polymer membrane using the DR Fast Spotter (DR. Chip Biotechnology Inc., Taiwan); four specific probes were used for each thrips species. After the spots dried, the microarray plate was put in a UV crosslinker to immobilize the probes. With 500 μl distilled water infused into each well for 5–10 min, 95% EtOH was added and then removed. The wells were allowed to dry at 45 °C.
Microarray hybridization and scanning
Microarrays were hybridized, washed and detected using the DR Chip DIY™ Kit. The spotted wells were immersed with 200 μl hybridization buffer. A 10 μl aliquot of the target PCR product generated from paired primers of 18Se and 5.8SThR for individual thrips, denatured at 94 °C for 5 min and then chilled on ice, was added to the well. The microarray was then incubated at 45 °C in the oven with vibration for 60 min. After removing the hybridization buffer, the well was washed by 250 μl wash buffer three times. The blocking solution, i.e., 0.2 μl Strep-AP mixed with 200 μl blocking reagent, was added in each well for 30 min, and then the well was washed with wash buffer three times. Detection solution, i.e., 4 μl NBT/BCIP mixed with 196 μl detection buffer, was added to the well for 5–10 min. The detection solution was then drawn away, and the well was washed with distilled water. The hybridized pattern was detected using DR. AIM™ reader (DR. Chip Biotechnology Inc., Taiwan).
Results
ITS1 sequence variation within and among thrips species
A total of 43 ITS1 sequences, ranging from 800 to 1250 bp, for 15 thrips species were obtained by PCR and have been deposited in GenBank (AB904169–AB904212). With ten sequences from GenBank for F. occidentalis, F. intonsa, and S. dorsalis included in the analysis, average sequence variation within species is less than 1%, except for those of S. dorsalis and T. palmi which are 11 and 3.5%, respectively. On the other hand, interspecific sequence distances were much higher than intraspecies distances, ranging from 15 to 56% (table 2). Deep phylogenetic divergences were found in thrips species, though there is a close relationship among Frankliniella species and between Thrips fuscipennis and Microcephalothrips abdominalis (data not shown).
Specificity and stability of specific primers
An examination of primer specificity and stability on 16 thrips species (table 1) shows the expected amplified products in target species with no cross amplifications (fig. 1). For each reaction, the successful generation of a PCR product of 520 bp by universal primers of 28S rDNA ensures a qualitative control for the entire experimental process. A few reactions with weak signals were likely due to competition or interference between primer pairs.
Fig. 1. Application of multiplex PCR by one ITS1 species-specific primer set of each thrips species with 28S rDNA universal paired primers. Specific amplification fragment is visible in target thrips with no cross amplification. Panels A and B: Dichromothrips smithi; C and D: Frankliniella cephalica; E and F: F. intonsa; G and H: F. occidentalis; I and J: Frankliniella williamsi; K and L: M. abdominalis; M and N: Megalurothrips usitatus; O and P: Stenchaetothrips biformis; Q and R: S. dorsalis; S and T: Thrips alliorum; U and V: T. florum; W and X: T. fuscipennis; Y and Z: T. hawaiiensis; a and b: T. palmi; c and d: T. tabaci. The first lane is 100 bp DNA ladder, and the examined thrips species are listed in Table 1.
DNA microarray for thrips identification
With two microarrays spotted with 11 and 15 thrips and four specific probes tested for each thrips species, all representative PCR products hybridized consistently to their corresponding probes. Probe Sbif3R, however, failed to detect Sbif PCR products (fig. 2), However, it showed no cross amplification in multiplex PCR (fig. 1). Probe Mabd3F showed weak signal in false hybridization to Fcep PCR products, and cross hybridization was observed for Tfus PCR product to probe Mabd (fig. 3).
Fig. 2. Hybridization pattern of microarray for 11 thrips species. The representative probes and spotted positions are shown in the right-bottom panel. Four spots in the corner are the positive controls and the hybrid signals of each target thrips are labeled with a box. Abbreviations for thrips species and their probes are as in Tables 1 and 2.
Fig. 3. Hybridization pattern of microarray for 15 thrips species. The representative probes and spotted positions are shown in the right-bottom panel. Two spots in the bottom-right corner box are the positive controls and hybrid signals of each target thrips are labeled with a box. Abbreviations for thrips species and their probes are as in Tables 1 and 2.
Discussion
The high variability of ITS1 suggests that it can be used for developing species-specific primers and probes for thrips identification. However, the sometimes low sequence divergence, e.g., between Mabd and Tfus and among Frankliniella species, may hamper species-specific primers designation. Fortunately, this examination of 16 thrips species confirms the specificity and stability of these primers (fig. 1). PCR-based identification for thrips species has been applied in a number of studies (Lin et al., Reference Lin, Wang and Yeh2003; Liu, Reference Liu2004; Rugman-Jones et al., Reference Rugman-Jones, Hoddle, Mound and Stouthamer2006; Asokan et al., Reference Asokan, Kumar, Kumar and Ranganath2007; Farris et al., Reference Farris, Ruiz-Arce, Ciomperlik, Vasquez and DeLeón2010; Huang et al., Reference Huang, Lee, Yeh, Shen, Mei and Chang2010; Kobayashi & Hasegawa, Reference Kobayashi and Hasegawa2012); however, none of them has introduced the internal control using universal primers pairs as we have done in this study, i.e., 28S rDNA, to ensure the DNA-template quality and optimal experimental procedures.
In the microarray assay, the weak signals shown by several hybridized spots might have resulted from an irregular manipulation of the spotted needles when spotting probes onto the polymer membrane. In the same microarray, some positive spots showed strong signal while others were weak or nearly invisible (fig. 3). Inconsistent signal intensity may have been due to the ΔT m values of probes (table 2). High sequence similarity between species might have increased the possibility of cross hybridization, as observed for target DNA of T. fuscipennis on the probes of M. abdominalis (fig. 3). Both phylogenetic relationships and sequence divergence (table 3) have revealed the close affinity between Mabd and Tfus. Regarding the possible mis-identification based on weak signal or cross hybridization, this study has adopted multiple probes for each thrips species in order to improve the accuracy of identification. In conclusion, we have demonstrated that multiplex PCR using universal primers with species-specific primers based on ITS1 sequences is a reliable, convenient and cost-effective diagnostic method to discriminate thrips species. Moreover, the microarray assay appears to be a comprehensive tool for the simultaneous identification of a number of thrips species. Of the ca. 6000 described species of thrips throughout the world, approximately 2% have been considered as crop pests (Inoue & Sakurai, Reference Inoue and Sakurai2007; Mound & Morris, Reference Mound and Morris2007), hence developing a high-throughput detection system with probes for the vast number of thrips pest in a single microarray is a worthwhile pursuit.
Table 3. Average sequence divergences between thrips species.
Acknowledgements
We thank Mr C.W. Chang and W.C. Tsao for collecting materials. We also like to thank Dr Ward Wheeler (American Museum of Natural History) to edit this manuscript. This work was supported by the National Science Council of Taiwan (NSC99-2324-B-005-020-CC2, NSC100-2324-B-005-013-CC2) and 99AS-9.3.1-BQ-B1 (Bureau of Animal and Plant Health Inspection and Quarantine, Taiwan). We also acknowledge the technical services provided by Sequencing Core Facility of the National Yang-Ming University Genome Research Center (YMGC), which is supported by National Science Council.
