Selection and evaluation of qPCR reference genes for expression analysis in the tiny egg parasitoid wasp, Trichogramma dendrolimi Matsumura (Hymenoptera: Trichogrammatidae)

The egg parasitoid Trichogramma spp. is an important biological control agent used against multiple species of Lepidopteran pest in forestry and agriculture. Due to the importance of Trichogramma spp. in biocontrol programs, its biological characteristics have been studied in detail, and current investigations should focus on the molecular biology of these tiny parasitoids. Real-time quantitative PCR (qPCR) is considered as the standard method for quantifying the gene expression of organisms. Surprisingly, the appropriate reference genes to ensure robust qPCR have not been documented at all for the Trichogramma genus. This study aimed to identify suitable reference genes for use in qPCR procedure of Trichogramma dendrolimi. Nine candidate housekeeping genes, namely glyceraldehyde-3-phosphate dehydrogenase (GAPDH), forkhead box O (FOXO), superoxide dismutase (SOD), beta-actin (ACTIN), ribosomal protein L10a (RPL10a), L18 (RPL18), L28 (RPL28), S13 (RPS13), and S15 (RPS15), were tested for their suitability as reference genes for developmental stage (3rd, 4th, 5th, 6th, 7th, 8th, 9th, and 10th day after parasitization), tissue (head, thorax, and abdomen of adults), sex of adults (male and female), and temperature (17 °C, 25 °C, and 32 °C). According to the GeNorm analysis, robust analysis should involve using an appropriate combination of reference genes, namely, at least three genes for different development stages, two genes for different tissues, two genes for different sex, and two genes for different temperature, respectively. According to the RelFinder method and by assessing the integrated values from using the ΔCt method, GeNorm, NormFinder, and BestKeeper, we identified the developmental stage-specific reference genes SOD, GAPDH, and ACTIN; tissue-specific reference genes RPL18 and RPS15; sex-specific reference genes SOD and RPL18; and temperature-specific reference genes RPL18 and RPL10. When testing the use of stable vs. unstable reference genes, the substantial differences were observed in the estimation expression of a hypothetical target gene, HSP90, in response to temperature. The present study provides a robust method for the measurement of gene expression in T. dendrolimi and will be helpful for future biological control programs using Trichogramma wasps.

characteristics have been studied in detail, and current investigations should focus on 23 the molecular biology of these tiny parasitoids. Real-time quantitative PCR (qPCR) is 24 considered as the standard method for quantifying the gene expression of organisms.

148
All the samples were quickly transferred into 1.5 mL tubes and stored at −80℃.

149
Each treatment or level involved three replicates.  BestKeeper.

188
GeNorm was applied to select the optimal reference gene by calculating a  NormFinder is used for ranking the constancy of candidate reference genes [17].

197
The calculation principle of NormFinder is similar to that of GeNorm. The optimal 198 reference gene was also selected based on a stable parameter of the candidate genes.

199
The criterion of NormFinder is the same as GeNorm.

205
The ∆Ct method was used to compare the stability of genes within the sample.

206
The ranks are based on the value of ∆Ct BestKeeper. According to RefFinder, the stability of genes among the developmental 243 stages was ranked as follows: (Fig. 2a). The GeNorm analysis suggested that all values 245 of Vn/Vn+1 were above 0.15. Therefore, three reference genes should be used when  (Table 3).  Table 2).

256
According to RefFinder, the stability of expression among tissues was ranked as 257 follows by gene: (Fig. 2b). According to results of GeNorm, the values of V2/3 were 259 below 0.15. Therefore, the combination of two reference genes is recommended for 260 adjusting the expression of genes in different tissues of T. dendrolimi (Fig. 3). The top 261 two stable genes, RPL18 and RPS15, formed the most appropriate combination of 262 reference genes for the adjustment of errors among tissues (Table 3).

264
According to the four algorithms, the top three stable genes between females and 265 males were SOD, RPL18, and RPL10a, and the top three unstable genes were FOXO, 266 RPL28, and RPS13 (Table 2).

267
According to RefFinder, the stability of expression as influenced by sex were influenced by sex (Fig. 3). SOD and RPL18 formed the most suitable combination of 273 reference genes for the adjustment of errors induced by sex (Table 3).  (Table 2).

280
According to RelFinder, the stability of expression as influenced by temperature 281 were ranked as follows by gene: (Fig. 2d). The results of GeNorm showed that the 283 values of V2/3 were below 0.15. Therefore, the combination of two reference genes 284 should be used for normalizing the expression of target genes as influenced by 285 temperature (Fig. 3). RPL18 and RPL10a formed the most suitable combination of 286 reference genes for the adjustment of errors induced by temperature (Table 3). was also no significant difference between 17 ℃ and 25 ℃ for results of HSP90 302 expression normalized by FOXO (Fig. 4).

304
The qPCR analysis is viewed as the model methodology to measure the the results were different to those adjusted using the recommended reference genes 312 (Fig. 4). This implies that the application of inappropriate reference genes reduces the 313 repeatability of experiments and lead to misleading conclusions.

314
In present study, four different algorithms, the ∆Ct method, NormFinder,

315
GeNorm, and BestKeeper, were used to identify the stability of nine candidate genes.

316
Although these algorithms produced globally similar rankings, some differences could 317 be observed due to the different criteria of algorithms. These discrepancies arise from 318 the specific working hypotheses and different premises of the algorithm. However, 319 the same general pattern could be observed across the different algorithms under most 320 conditions.

321
In our study, the RPL18 gene was selected as a tissue-specific,   Tukey-Kramer test (P < 0.05).