- Email: [email protected]
Selection of reference genes for quantitative real-time PCR analysis in chicken ovary following silver nanoparticle treatment
Environmental Toxicology and Pharmacology 56 (2017) 186–190
Contents lists available at ScienceDirect
Environmental Toxicology and Pharmacology journal homepage: www.elsevier.com/locate/etap
Research Paper
Selection of reference genes for quantitative real-time PCR analysis in chicken ovary following silver nanoparticle treatment
MARK
⁎
Dorota Katarzyńska-Banasik , Małgorzata Grzesiak, Andrzej Sechman Department of Animal Physiology and Endocrinology, University of Agriculture in Krakow, Al. Mickiewicza 24/28, 30-059 Krakow, Poland
A R T I C L E I N F O
A B S T R A C T
Keywords: Silver nanoparticles Reference gene Chicken Ovary
Reliable results of quantitative real time PCR (qPCR) analysis require normalization of target gene expression level using reference genes (RGs). However housekeeping genes expression may vary under experimental conditions, so selection of the proper RGs is a crucial step in a qPCR analysis. Several algorithms have been developed to address this problem: geNorm, NormFinder and BestKeeper. In this study, we have used these three tools to evaluate the stability of RGs in the ovarian tissues of hens treated with silver nanoparticles. Eight genes were selected for the validation: HPRT, HMBS, VIM, SDHA, TBP, RPL13, GAPDH and 18S rRNA. According to geNorm the best combination of reference genes is SDHA and TPP. NormFinder also selected SDHA as the most suitable gene, but in combination with RPL13. Analysis in BestKeeper showed that SDHA, RPL13 might be the best choice in gene expression studies using the chicken ovary. In conclusion, the results obtained depend on the algorithm used and it arises from the diverse calculation strategies used in these programs. The outcome from the NormFinder is considered to be the most trustworthy and used in further qPCR analysis.
1. Introduction Real time PCR is currently a leading method used in determination of gene expression (Bustin et al., 2005; Wong and Medrano, 2005). There are some difficulties connected with non-biological variations that need to be overcome in gene expression analysis e.g. sample preparation and storage, poor reverse transcription, mRNA quality, and cDNA concentration (Bustin, 2002; Lekanne Deprez et al., 2002; Sanders et al., 2014). Several different strategies of normalization are used to correct these sample-to-sample and run-to-run variations, but the most preferred way is normalization against reference gene/s (RG/ s) (Huggett et al., 2005). Perfect RGs should be unaffected by experimental conditions and show minimal variation between tissues. It seems that the best candidates to meet these criteria are ubiquitously expressed housekeeping genes (HKGs). Over the years, the most commonly used HKGs have been β-actin (ACTB), glyceraldeyde-3-phosphate dehydrogenase (GAPDH), ribosome small subunit ribosomal RNA (18S rRNA), and 28S ribosomal rRNA (28S rRNA) (Suzuki et al., 2000). On the other hand, numerous studies argue that they are unsuitable internal controls with variable expression (Glare et al., 2002; Ohl et al., 2005; Selvey et al., 2001; Thellin et al., 1999; Toegel et al., 2007). Since it has become clear that a universal RG, one constitutively expressed regardless of conditions, does not exist, validation of RG stability in each experiment is highly required to obtain a reliable expression
⁎
analysis (Kozera and Rapacz, 2013; Radonic et al., 2004). The current approach to analysis of gene expression also demands the use of more than one RG (Bustin et al., 2009; Thellin et al., 2009) to preclude considerably large bias (Derveaux et al., 2010; Nicot et al., 2005). According to Vandesompele et al. (2002) quantification analysis using validated multiple internal controls gives the most appropriate results. The two most important algorithms for determination of the stability of RGs are geNorm and NormFinder. GeNorm creates stability ranking of tested RGs and specifies the minimum number of genes for accurate normalization of gene expression (Vandesompele et al., 2002). On the other hand, NormFinder calculates intra- and inter-group variation as well as creates a ranking of the most stably expressed genes (Andersen et al., 2004). There is another well-known statistical tool defining the variability of RGs called BestKeeper, which creates the descriptive statistics with several measures and leaves the choice of selection parameters to the experimenter (Pfaffl et al., 2004). Literature regarding the selection of RGs in chicken tissues is scarce. Existing reports are related to infectious diseases (Kuchipudi et al., 2012; Yang et al., 2013; Yin et al., 2011; Yue et al., 2010), immunological research (Borowska et al., 2016; De Boever et al., 2008; Mitra et al., 2016) or different tissues (Bagés et al., 2015; Nascimento et al., 2015), while there are no reports associated with toxicological studies. In this type of research the selection of a proper internal control gene is extremely important due to the significant impact of
Corresponding author. E-mail address: [email protected] (D. Katarzyńska-Banasik).
http://dx.doi.org/10.1016/j.etap.2017.09.011 Received 5 June 2017; Received in revised form 17 September 2017; Accepted 18 September 2017 Available online 20 September 2017 1382-6689/ © 2017 Published by Elsevier B.V.
Environmental Toxicology and Pharmacology 56 (2017) 186–190
D. Katarzyńska-Banasik et al.
Table 1 The summary of TaqMan Gene Expression Assays characteristic. Gene symbol
Assay ID
Context sequence
Amplicon size (bp)
SDHA RPL13 TBP HMBS VIM GAPDH HPRT1 CYP19A1 18S rRNA
Gg03330765_m1 Gg03348054_m1 Gg03366488_m1 Gg03317022_m1 Gg03360311_m1 Gg03346982_m1 Gg03338899_m1 Gg03346001_m1 Eukaryotic 18S rRNA Endogenous Control
GCAGAAGACAATGCAAAGCCATGCT TTATGCCGATCAGGAACGTTTTCAA CAGGAGCAAAAAGCGAGGAACAGTC GCTTTGAGATTGTTGCCATGTCTAC GGAAACTAGAGATGGACAGGTTATT TCGTCAAGCTTGTTTCCTGGTATGA GATATCGGCCAGACTTTGTTGGATT ATTGAAACTGTTATGGGTGACAGAG
103 66 82 87 85 107 89 76 187
SDHA – Succinate dehydrogenase complex flavoprotein subunit A, RPL13 – 60S ribosomal protein L13, TBP – TATA-binding protein, HMBS – Hydroxymethylbilane synthase, VIM –vimentin, GAPDH – Glyceraldehyde 3-phosphate dehydrogenase, HPRT1– Hypoxanthine-guanine phosphoribosyltransferase, CYP19A1 – Cytochrome P450, family 19, subfamily a, polypeptide 1, 18S rRNA – 18S ribosomal RNA.
Transcription Kit (Thermo Scientific) following manufacturer's recommendations. First strand cDNA was synthesized using random primers for 5 min at 25 °C followed by 60 min at 42 °C; reaction was terminated by heating at 70 °C for 5 min. Quantitative real time PCR (qPCR) was performed on 96-well plates in a StepOne Plus thermocycler (Applied Biosystems, Foster City, CA, USA). Each reaction was performed in duplicate in a total volume of 10 μl. One ul of 10-fold diluted cDNA was added to 9 μl of reaction mix containing 5 μl TaqMan Gene Expression Master Mix (Applied Biosystems), 3.5 μl nuclease free water, and 0.5 μl TaqMan Gene Expression Assay (Applied Biosystems). Thermal cycling conditions were as follows: 2 min at 50 °C, 10 min at 95 °C and 40 cycles of denaturation for 15 s at 95 °C, and 1 min of annealing at 60 °C. Standard curves from 10-fold serial dilutions of pooled cDNA were generated to determine amplification efficiency for each tested gene. Amplification efficiencies were as follows: HPRT – 97%, HMBS – 96%, VIM – 99%, 18S rRNA – 95%, GAPDH – 100%, SDHA – 95%, RPL13–102%, TBP – 99%. No template controls with nuclease free water were prepared in each plate in order to check reagent contamination. TaqMan Gene Expression Assays containing primers and probe were chosen from predesigned assays offered by Applied Biosystems (Table 1). Designed probes spanned exon–exon junctions in order to avoid amplification of genomic DNA. Moreover, we used minus-reverse transcriptase controls to assesses the amount of DNA contamination for each gene. 2% agarose gel electrophoresis was prepared in order to verify the amplicons size of each tested gene and it revealed that all gene expression assays amplified a single PCR product with the expected size (Fig. 1). Statistical analysis was carried out using three available software programs: geNorm, NormFinder and BestKeeper. GeNorm calculates the gene expression stability value (M) for the reference gene as the average pairwise variation V for that gene with all other tested reference genes. It creates a ranking of stable RGs by stepwise elimination of genes with the highest M value, determining the two best RGs. GeNorm also indicates the minimum number of genes that should be used for normalization (Vandesompele et al., 2002). This algorithm is sensitive to co-regulation so selected genes must stem from different biological pathways (Andersen et al., 2004). The input data for geNorm were the quantities obtained through transformation of Ct values according to the delta Ct method (Q = E(minCt − sampleCt)). NormFinder is a model based tool, which calculates the M of RGs taking into account intra and inter-group variations. It also suggests the two most stable genes and provides their combined stability value. This model is considered less susceptible to co-regulation (Andersen et al., 2004). Analysis in NormFinder was made on Ct values converted to linear scale according to formula E−ΔCt and then log-transformed by the program itself. BestKeeper creates descriptive statistics with multiple factors, i.e. the geometric mean (GM), arithmetic mean (AM), minimal (Min) and maximal (Max) values, standard deviation (SD), and coefficient of
experimental factors (Arukwe, 2006). The increasing application of silver nanoparticles (AgNPs) in the contemporary world creates the necessity to verify their safety for humans and the environment. AgNPs are now considered as new additives in poultry nutrition and used as disinfectants and odorant limiting agents (Farzinpour and Karashi, 2013; Pineda et al., 2012). In females of domestic birds the number of eggs produced, reproductive rates and breeding results depend on the number of oocytes produced in the ovary, and any anomalies in this area are highly undesirable by breeders. Therefore, the aim of this study was to select the best RGs using normalization of gene expression in the ovaries of hens treated with AgNPs. Eight candidate RGs were analyzed: hypoxanthine-guanine phosphoribosyltransferase (HPRT), hydroxymethylbilane synthase (HMBS), vimentin (VIM), succinate dehydrogenase complex flavoprotein subunit A (SDHA), TATA-binding protein (TBP), 60S ribosomal protein L13 (RPL13), GAPDH, and 18S rRNA. Their stability was assessed using geNorm, NormFinder and BestKeeper software tools.
2. Materials and methods All procedures were performed with the permission of the Local Animal Ethics Committee in Krakow (no. 9/2015). Hy-Line Brown chickens at the age of 25 weeks were kept in individual cages with free access to water and feed, on a 14L:10D lighting schedule. Chickens received per os 1 ml/kg b.w. of colloidal AgNPs in two sizes: 13 nm and 50 nm and in concentrations of 10 ppm or 100 ppm, respectively. The control group was treated with the reference solution in which AgNPs were suspended. After 14-day administration of AgNPs, the chickens were decapitated and prehierachical and hierarchical ovarian follicles were obtained. Granulosa and theca layers were separated from hierarchical ovarian follicles and stored separately. Tissues were snapfrozen in liquid nitrogen immediately after isolation (c.a ∼2–3 min) and stored at −80 °C up to 3 months until RNA extraction from all tissues. All collected samples of ovarian tissues were analyzed together to select the best reference gene for the whole ovary; i.e. from each treatment group (13 nm, 50 nm, control group) we randomly chose 4 samples from prehierachical follicles, 5 granulosa layers from F2 follicles and 5 theca layers from F2 follicles (i.e. 14 biological replicates in each treatment group). It means, that 42 (i.e. 14 replicates ×3 groups) biological repetitions were used in the whole reference gene selection analysis. Total RNA was isolated using silica membrane technology with commercial kit Extractme Total RNA according to the manufacturer's instructions (DNA-Gdańsk, Poland). RNA quality and quantity were measured using NanoDrop Lite (Thermo Scientific, Madison, USA) and checked on 1% agarose gel electrophoresis. The A260/A280 ratios were quantified to assess the quality of the extracted RNA, for all analyzed samples were between 1.96 and 2. Two μg of total RNA was reverse transcribed using a RevertAid RT 187
Environmental Toxicology and Pharmacology 56 (2017) 186–190
D. Katarzyńska-Banasik et al.
Fig. 1. Agarose gel (2%) showing amplification of a specific PCR product of the expected size for the tested genes.
variance (CV), and determines the BestKeeper index (BI). Finally BI is correlated with each RG providing Pearson coefficient correlation (r), coefficient of determination (r2) and p-value. There is no ranking of RGs in BestKeeper and it requires row Ct values to perform the analysis (Pfaffl et al., 2004). CYP19A1 gene, which is one of the steroidogenic genes, was selected to evaluate the reliability of tested reference genes. Amplification efficiency of this gene was 101%. Normalized relative quantity (NRQ) of CYP19A1 gene was calculated using Pfaffl (2001) method and Hellemans et al. (2007) extended model.
0.15
0.1
0.05
3. Results 0
3.1. geNorm Fig. 3. Pairwise variation (Vn/Vn + 1) generated by geNorm for determination of optimal number of reference genes (n = 42).
SDHA and TBP are the most stable reference genes according to geNorm, yielding a combined stability value of M = 0.49 (Fig. 2). Besides geNorm determines the optimal number of RGs by calculating the pairwise variation Vn/Vn+1 between two sequential normalization factors NFn and NFn+1 that contain an increasing number of RGs. Since Vn/Vn+1 recommended threshold is 1.5, this analysis suggests using at least 3 reference genes (V3/4 < 0.15) (Fig. 3). Consequently SDHA, TBP and GAPDH are suggested to be used for normalization as they exhibit the lowest M values. 3.2. NormFinder According to NormFinder the most stable genes are SDHA (M = 0.076) and RPL13 (M = 0.089) (Fig. 4) and their combined stability value is 0.06. Analysis of intra- and inter-group variations for each group are shown in Fig. 5. The columns indicate the inter-group variation and the error bars represent the average of the intra-group variances. This graph clearly demonstrates that the best genes are those with a minimal inter-group variation and the smallest possible error bars.
Fig. 4. Reference genes stability values evaluated by NormFinder (n = 42).
3.3. BestKeeper Descriptive statistics are shown in Table 2. Taking into account SD values HPRT (SD = 0.58), VIM (SD = 0.61), RPL13 (SD = 0.63) and 18S rRNA (SD = 0.75) are the most stable genes, while GAPDH is the worst one (SD = 1.08). However based on correlation coefficient (r), the ranking is as follows: SDHA (r = 0.894), TBP (r = 0.885), GAPDH (r = 0.865), RPL13 (r = 0.854); and 18S rRNA is characterized by the lowest correlation coefficient (r = 0.635). When we combine these two variables, as described in the discussion, SDHA seem to be the best choice, then RPL13 and TBP.
0.90
0.9
0.82
0.8
0.77 0.7
0.68 0.61
0.6
0.54 0.5 0.4 18S
VIM
HPRT
RPL13
HMBS
GAPDH
3.4. Reference gene validation
SDHA TBP
CYP19A1 gene expression was calculated by using all tested genes for normalization (Fig. 6). When RPL13 and SDHA were used for normalization, the expression levels of CYP19A1 gene were equal in
Fig. 2. Average expression stability values of reference genes evaluated by geNorm (n = 42).
188
Environmental Toxicology and Pharmacology 56 (2017) 186–190
D. Katarzyńska-Banasik et al.
greater (it decreased by more than 80% in 50 nm group). The same difference in NRQs was observed when genes selected by geNorm were used for normalization, i.e. SDHA and TBP or SDHA, TBP and GAPDH. Normalization against the most unstable gene 18S rRNA lead to elevated levels of CYP19A1 gene expression in both groups 13 and 50 nm AgNPs.
0.40 0.20 Control
0.00 -0.20 Intr
4. Discussion
-0.40
The necessity to normalize data obtained from real-time PCR toward stable validated reference genes (RGs) is indisputable, especially in toxicological studies. Thus far, we have not found information about the appropriate endogenous control for research conducted on the chicken ovary, nor any toxicological studies in relation to this organ. Olias et al. (2014) evaluated RGs in the ovarian tissue of song birds (such as zebra finches), but there are no reports on the validation of RGs in chicken tissues, where the influence of toxic agents has been investigated. Except for genes selected on the basis of research conducted on zebra finches (Olias et al., 2014), we have included two commonly used internal controls: GAPDH and 18S rRNA. These genes served as individual RGs for normalization of qPCR data in multiple studies in chickens (Adams et al., 2009; Hong et al., 2006; Hrabia et al., 2013; Richards et al., 2006; Zhang et al., 2012). In our study, 18S rRNA was revealed to be the least stable gene by geNorm and NormFinder. In BestKeeper, 18S rRNA had the worst coefficient correlation, while GAPDH was characterized by the highest SD. Analysis conducted separately in control and each experimental group confirmed that the 18S rRNA gene is highly unstable (data not shown). That is in agreement with the previous findings in avian gonads (Olias et al., 2014). Generally, the 18S rRNA gene has some disadvantages and in many cases is not recommended for normalization (Kozera and Rapacz, 2013). Results regarding the most stable expressed genes were similar in BestKeeeper and NormFinder analyses, but entirely different in geNorm. BestKeeper does not specify which genes are the most stable and the experimenter has to decide for himself which parameter is the most meaningful. Although the combination of SD and the correlation coefficient of BI with each RG is predominantly used (Olias et al., 2014; Wood et al., 2008; Normann et al., 2016), in some studies only one of these variables is taken into account: only SD (Wang et al., 2012; Ostrowska et al., 2014) or only r (Maroufi et al., 2010). We decided to combine SD and coefficient correlation. Following the suggestion of Pfaffl et al. (2004) we excluded genes with SD higher than 1, such as GAPDH. Then we decided to rank genes based on SD and coefficient of correlation higher than 0.8. Implementing this method SDHA, RPL13 and TBP were considered as the most stably expressed genes. The same sequence of genes was indicated by NormFinder, which suggested that SDHA and RPL13 together were the best combination of genes for qPCR data normalization. Another pair of genes was selected by geNorm, namely the SDHA and TPB. The reason for these differences are diverse calculation strategies used in these programs. GeNorm uses a pairwise comparison approach and selects two genes with lowest intragroup variation and close enough intergroup variation. Consequently, the selected genes will have similar expression profiles but they will not necessarily be the most stable genes (Andersen et al., 2004). On the other hand, NormFinder selects two candidates with minimal shared inter- and intragroup variation in a so-called modelbased approach (Andersen et al., 2004). The model-based strategy seems to be more precise and trustworthy, so we decided to make use of SDHA and RPL13 as the most stable genes identified by both BestKeeper and NormFinder for further analysis of gene expression. Our choice was confirmed by CYP19A1 gene expression calculations and the analysis showed that normalization against only one of the selected genes SDHA or RPL13 would not generate different expression patterns than normalization against both of them. CYP19A1 gene expression also revealed that normalization using the genes selected by GeNorm (SDHA, TBP) yields other expression results than normalization with respect to
-0.60
Fig. 5. Inter-and intra-group variation evaluated by NormFinder. The graph bars and the error bars represent the inter-group variation and an average of the intra-group variation, respectively (n = 14 in each group).
Table 2 BestKeeper descriptive statistics of the analyzed genes in the chicken ovary. Statistics derived from the 42 biological replicates.
n gmean SD [ ± Ct] CV [%Ct] Bk [r] p-value
HPRT
VIM
RPL13
18S
HMBS
SDHA
TBP
GAPDH
42 26.31 0.58 2.19 0.737 0.001
42 24.49 0.61 2.51 0.667 0.001
42 25.11 0.63 2.53 0.854 0.001
42 13.14 0.75 5.69 0.635 0.001
42 27.00 0.85 3.14 0.796 0.001
42 28.63 0.86 3.01 0.894 0.001
42 28.63 0.97 3.38 0.885 0.001
42 25.03 1.08 4.31 0.865 0.001
SDHA – Succinate dehydrogenase complex flavoprotein subunit A, RPL13 – 60S ribosomal protein L13, TBP – TATA-binding protein, HMBS – Hydroxymethylbilane synthase, VIM – vimentin, GAPDH – Glyceraldehyde 3-phosphate dehydrogenase, HPRT1– Hypoxanthineguanine phosphoribosyltransferase, 18S rRNA – 18S ribosomal RNA, gmean – geometric mean, SD [ ± Ct] – standard deviation of the Ct, CV [%Ct] – the coefficient of variance expressed as a percentage of the Ct level, Bk [r] – Pearson’s coefficient correlation of each gene to the BI-index, p-value – probability value.
Fig. 6. Relative quantification of CYP19A1 gene expression using tested reference genes for normalization under AgNPs treatment in chicken ovary. Each value represents the mean ± SEM from five determinations.
control and 13 nm AgNPs group and decreased in 50 nm AgNPs group by approximately 30%. The same expression patterns were observed when RPL13 and SDHA were used separately. When less stable reference genes TBP, HMBS, GAPDH, HPRT and VIM were used for normalization the expression levels increased in 13 nm AgNPs group and the difference between NRQ in 13 nm group and 50 nm group was 189
Environmental Toxicology and Pharmacology 56 (2017) 186–190
D. Katarzyńska-Banasik et al.
Lekanne Deprez, R.H., Fijnvandraat, A.C., Ruijter, J.M., Moorman, A.F.M., 2002. Sensitivity and accuracy of quantitative real time polymerase chain reaction using SYBR green I depends on cDNA synthesis conditions. Anal. Biochem. 307, 63–69. Maroufi, A., Van Bockstaele, E., De Loose, M., 2010. Validation of reference genes for gene expression analysis in chicory (Cichorium intybus) using quantitive real-time PCR. BMC Mol. Biol. 11, 1–12. Mitra, T., Bilic, I., Hess, M., Liebhart, D., 2016. The 60S ribosomal protein L13 is the most preferable reference gene to investigate gene expression in selected organs from turkeys and chickens, in context of different infection models. Vet. Res. 47, 105. Nascimento, C.S., Barbosa, L.T., Brito, C., Fernandes, R.P.M., Mann, R.S., Pinto, A.P., Oliveira, H.C., Dodson, M.V., Guimarães, S.E.F., Duarte, M.S., 2015. Identification of suitable reference genes for real time quantitative polymerase chain reaction assays on Pectoralis major muscle in chicken (Gallus gallus). PLoS One 10 (5), e0127935. Nicot, N., Hausman, J.F., Hoffmann, L., Evers, D., 2005. Housekeeping gene selection for real-time RT-PCR normalization in potato during biotic and abiotic stress. J. Exp. Bot. 56, 2907–2914. Normann, K.R., Øystese, K.A., Berg, J.P., Lekva, T., Berg-Johnsen, J., Bollerslev, J., Olarescu, N.C., 2016. Selection and validation of reliable reference genes for RTqPCR analysis in a large cohort of pituitary adenomas. Mol. Cell. Endocrinol. 437, 183–189. Ohl, F., Jung, M., Xu, C., Stephan, C., Rabien, A., Burkhardt, M., Nitsche, A., Kristiansen, G., Loening, S.A., Radonic, A., Jung, K., 2005. Gene expression studies in prostate cancer tissue: which reference gene should be selected for normalization? J. Mol. Med. 83, 1014–1024. Olias, P., Adam, I., Meyer, A., Scharff, C., Gruber, A.D., 2014. Reference genes for quantitative gene expression studies in multiple avian species. PLoS One 9, e99678. Ostrowska, M., Żelazowska, B., Słoniewski, K., Kowalski, Z.M., Zwierzchowski, L., 2014. Technical note: selecting the best references in gene expression experiments in liver of cows receiving glucogenic supplements during the transition period. J. Dairy Sci. 97, 911–916. Pfaffl, M.W., Tichopad, A., Prgomet, C., Neuvians, T.P., 2004. Determination of stable housekeeping genes, differentially regulated target genes and sample integrity: BestKeeper–excel-based tool using pair-wise correlations. Biotechnol. Lett. 26, 509–515. Pfaffl, M.W., 2001. A new mathematical model for relative quantification in real-time RTPCR. Nucleic Acids Res. 29 (9), e45. Pineda, L., Chwalibog, A., Sawosz, E., Lauridsen, C., Engberg, R., Elnif, J., Hotowy, A., Sawosz, F., Gao, Y., Ali, A., Moghaddam, H.S., 2012. Effects of silver nanoparticles on growth performance, metabolism and microbial profile of broiler chickens. Arch. Anim. Nutr. 66, 416–429. Radonic, A., Thulke, S., Mackay, I.M., Landt, O., Siegert, W., Nitsche, A., 2004. Guideline to reference gene selection for quantitative real-time PCR. Biochem Biophys Res. Commun. 313, 856–862. Richards, M.P., Poch, S.M., McMurtry, J.P., 2006. Characterization of turkey and chicken ghrelin genes, and regulation of ghrelin and ghrelin receptor mRNA levels in broiler chickens. Gen. Comp. Endocrinol. 145, 298–310. Sanders, R., Mason, D., Foy, C., Huggett, J., 2014. Considerations for accurate gene expression measurement by reverse transcription quantitative PCR when analysing clinical samples. Anal. Bioanal. Chem. 406, 6471–6483. Selvey, S., Thompson, E.W., Matthaei, K., Le, A.R.A., Irving, M.G., Griffiths, L.R., 2001. Beta-actin—an unsuitable internal control for RT-PCR. Mol. Cell. Probes 15, 307–311. Suzuki, T., Higgins, P.J., Crawford, D.R., 2000. Control selection for RNA quantitation. Biotechniques 29, 332–337. Thellin, O., Zorzi, W., Lakaye, B., De Borman, B., Coumans, B., Hennen, G., Grisar, T., Igout, A., Heinen, E., 1999. Housekeeping genes as internal standards: use and limits. J. Biotechnol. 75, 291–295. Thellin, O., ElMoualij, B., Heinen, E., Zorzi, W., 2009. A decade of improvements in quantification of gene expression and internal standard selection. Biotechnol. Adv. 27, 323–333. Toegel, S., Huang, W., Piana, C., Unger, F.M., Wirth, M., Goldring, M.B., Gabor, F., Viernstein, H., 2007. Selection of reliable reference genes for qPCR studies on chondroprotective action. BMC Mol. Biol. 8, 13. Vandesompele, J., De Preter, K., Pattyn, F., Poppe, B., Van Roy, N., De Paepe, A., Speleman, F., 2002. Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol. 3 (research0034):1. Wang, Q., Ishikawa, T., Michiue, T., Zhu, B.L., Guan, D.W., Maeda, H., 2012. Stability of endogenous reference genes in postmortem human brains for normalization of quantitative real-time PCR data: comprehensive evaluation using geNorm NormFinder, and BestKeeper. Int. J. Leg. Med. 126, 943–952. Wong, M.L., Medrano, J.F., 2005. Real time PCR for mRNA quantification. Biotechniques 39, 75–85. Wood, S.H., Clements, D.N., McEwan, N.A., Nuttall, T., Carter, S.D., 2008. Reference genes for canine skin when using quantitative real-time PCR. Vet. Immunol. Immunopathol. 126, 392–395. Yang, F., Lei, X., Rodriguez-Palacios, A., Tang, C., Yue, H., 2013. Selection of reference genes for quantitative real-time PCR analysis in chicken embryo fibroblasts infected with avian leukosis virus subgroup. J. BMC Res. Notes 6, 402. Yin, R., Liu, X., Liu, C., Ding, Z., Zhang, X., Tian, F., Liu, W., Yu, J., Li, L., Hrabé de Angelis, M., Stoeger, T., 2011. Systematic selection of housekeeping genes for gene expression normalization in chicken embryo fibroblasts infected with Newcastle disease virus. Biochem. Biophys. Res. Comm. 413, 537–540. Yue, H., Lei, X.W., Yang, F.L., Li, M.Y., Tang, C., 2010. Reference gene selection for normalization of PCR analysis in chicken embryo fibroblast infected with H5N1 AIV. Virol Sin. 25, 425–431. Zhang, W., Sargis, R.M., Volden, P.A., Carmean, C.M., Sun, X.J., Brady, M.J., 2012. PCB 126 and other dioxin-like PCBs specifically suppress hepatic PEPCK expression via the aryl hydrocarbon receptor. PLoS One 7 (5), e37103.
the genes recommended by NormFinder. It is not surprising because SDHA and TBP showed the same nonvanishing intergroup variation. 5. Conclusions In summary, the stability of RGs changes under different experimental conditions. This is the first report undertaken to select RGs for silver nanoparticle toxicity studies in the chicken ovary. In this study, we have determined that 18S rRNA is not stable in the chicken ovarian tissues and we do not recommend utilizing it in such studies. From the eight analyzed genes, we have chosen SDHA and RPL13 for further qPCR analysis. Conflict of interest The authors declare that there is no conflict of interest regarding the publication of this paper. Acknowledgments This study was supported by the National Science Centre, Poland [grant number UMO-2014/15/N/NZ9/01435]. The authors thank Professor Anna Hrabia and Dr Maria Mika for technical help during the experiment, Ms. Kinga Kowalik for assisting in RNA isolation and Mrs Balbina Sobesto for animal care. References Adams, S.C., Xing, Z., Li, J., Cardona, C.J., 2009. Immune-related gene expression in response to H11N9 low pathogenic avian influenza virus infection in chicken and Pekin duck peripheral blood mononuclear cells. Mol. Immunol. 46, 1744–1749. Andersen, C.L., Jensen, J.L., Ørntoft, T.F., 2004. 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 Res. 64, 5245–5250. Arukwe, A., 2006. Toxicological housekeeping genes: do they really keep the house? Environ. Sci. Technol. 40, 7944–7949. Bagés, S., Estany, J., Tor, M., Pena, R.N., 2015. Investigating reference genes for quantitative real-time PCR analysis across four chicken tissues. Gene 561, 82–87. Borowska, D., Rothwell, L., Bailey, R.A., Watson, K., Kaiser, P., 2016. Identification of stable reference genes for quantitative PCR in cells derived from chicken lymphoid organs. Vet. Immunol. Immunopathol. 170, 20–24. Bustin, S.A., Benes, V., Nolan, T., Pfaffl, M.W., 2005. Quantitative real-time RT-PCR–a perspective. J. Mol. Endocrinol. 34, 597–601. Bustin, S.A., Benes, V., Garson, J.A., Hellemans, J., Huggett, J., Kubista, M., Mueller, R., Nolan, T., Pfaffl, M.W., Shipley, G.L., Vandesompele, J., Wittwer, C.T., 2009. The MIQE guidelines: minimum information for publication of quantitative real-time PCR experiments. Clin. Chem. 55, 611–622. Bustin, S.A., 2002. Quantification of mRNA using real time reverse transcription PCR (RTPCR): trends and problems. J. Mol. Endocrinol. 29, 23–39. De Boever, S., Vangestel, C., De Backer, P., Croubels, S., Sys, S.U., 2008. Identification and validation of housekeeping genes as internal control for gene expression in an intravenous LPS inflammation model in chickens. Vet. Immunol. Immunopathol. 122, 312–317. Derveaux, S., Vandesompele, J., Hellemans, J., 2010. How to do successful gene expression analysis using real-time PCR. Methods 50, 227–230. Farzinpour, A., Karashi, N., 2013. The effects of silver nanoparticles on egg quality traits in laying Japanese quail. Appl Nanosci. 3, 95–99. Glare, E.M., Divjak, M., Bailey, M.J., Walters, E.H., 2002. Beta-actin and GAPDH housekeeping gene expression in asthmatic airways is variable and not suitable for normalising mRNA levels. Thorax 57, 765–770. Hellemans, J., Mortier, G., De Paepe, A., Spelman, F., Vandesompele, J., 2007. qBase relative quantification framework and software for management and automated analysis of real-time quantitative PCR data. Genome Biol. 8 (2), R19. Hong, Y.H., Lillehoj, H.S., Lillehoj, E.P., Lee, S.H., 2006. Changes in immune-related gene expression and intestinal lymphocyte subpopulations following Eimeria maxima infection of chickens. Vet. Immunol. Immunopathol. 114, 259–272. Hrabia, A., Leśniak, A., Sechman, A., 2013. In vitro effects of TCDD, PCB126 and PCB153 on estrogen receptors, caspases and metalloproteinase-2 mRNA expression in the chicken shell gland. Folia Biol. 61, 3–4. Huggett, Y., Dheda, K., Bustin, S., Zumala, A., 2005. Real-time RT-PCR normalization; strategies and considerations. Genes Immun. 6, 279–284. Kozera, B., Rapacz, M., 2013. Reference genes in real-time PCR. J. Appl. Genet. 54, 391–406. Kuchipudi, S.V., Tellabati, M., Nelli, R.K., White, G.A., Perez, B.B., Sebastian, S., Slomka, M.J., Brookes, S.M., Brown, I.H., Dunham, S.P., 2012. 18S rRNA is a reliable normalisation gene for real time PCR based on influenza virus infected cells. Virol. J. 9, 230.
190
Contents lists available at ScienceDirect
Environmental Toxicology and Pharmacology journal homepage: www.elsevier.com/locate/etap
Research Paper
Selection of reference genes for quantitative real-time PCR analysis in chicken ovary following silver nanoparticle treatment
MARK
⁎
Dorota Katarzyńska-Banasik , Małgorzata Grzesiak, Andrzej Sechman Department of Animal Physiology and Endocrinology, University of Agriculture in Krakow, Al. Mickiewicza 24/28, 30-059 Krakow, Poland
A R T I C L E I N F O
A B S T R A C T
Keywords: Silver nanoparticles Reference gene Chicken Ovary
Reliable results of quantitative real time PCR (qPCR) analysis require normalization of target gene expression level using reference genes (RGs). However housekeeping genes expression may vary under experimental conditions, so selection of the proper RGs is a crucial step in a qPCR analysis. Several algorithms have been developed to address this problem: geNorm, NormFinder and BestKeeper. In this study, we have used these three tools to evaluate the stability of RGs in the ovarian tissues of hens treated with silver nanoparticles. Eight genes were selected for the validation: HPRT, HMBS, VIM, SDHA, TBP, RPL13, GAPDH and 18S rRNA. According to geNorm the best combination of reference genes is SDHA and TPP. NormFinder also selected SDHA as the most suitable gene, but in combination with RPL13. Analysis in BestKeeper showed that SDHA, RPL13 might be the best choice in gene expression studies using the chicken ovary. In conclusion, the results obtained depend on the algorithm used and it arises from the diverse calculation strategies used in these programs. The outcome from the NormFinder is considered to be the most trustworthy and used in further qPCR analysis.
1. Introduction Real time PCR is currently a leading method used in determination of gene expression (Bustin et al., 2005; Wong and Medrano, 2005). There are some difficulties connected with non-biological variations that need to be overcome in gene expression analysis e.g. sample preparation and storage, poor reverse transcription, mRNA quality, and cDNA concentration (Bustin, 2002; Lekanne Deprez et al., 2002; Sanders et al., 2014). Several different strategies of normalization are used to correct these sample-to-sample and run-to-run variations, but the most preferred way is normalization against reference gene/s (RG/ s) (Huggett et al., 2005). Perfect RGs should be unaffected by experimental conditions and show minimal variation between tissues. It seems that the best candidates to meet these criteria are ubiquitously expressed housekeeping genes (HKGs). Over the years, the most commonly used HKGs have been β-actin (ACTB), glyceraldeyde-3-phosphate dehydrogenase (GAPDH), ribosome small subunit ribosomal RNA (18S rRNA), and 28S ribosomal rRNA (28S rRNA) (Suzuki et al., 2000). On the other hand, numerous studies argue that they are unsuitable internal controls with variable expression (Glare et al., 2002; Ohl et al., 2005; Selvey et al., 2001; Thellin et al., 1999; Toegel et al., 2007). Since it has become clear that a universal RG, one constitutively expressed regardless of conditions, does not exist, validation of RG stability in each experiment is highly required to obtain a reliable expression
⁎
analysis (Kozera and Rapacz, 2013; Radonic et al., 2004). The current approach to analysis of gene expression also demands the use of more than one RG (Bustin et al., 2009; Thellin et al., 2009) to preclude considerably large bias (Derveaux et al., 2010; Nicot et al., 2005). According to Vandesompele et al. (2002) quantification analysis using validated multiple internal controls gives the most appropriate results. The two most important algorithms for determination of the stability of RGs are geNorm and NormFinder. GeNorm creates stability ranking of tested RGs and specifies the minimum number of genes for accurate normalization of gene expression (Vandesompele et al., 2002). On the other hand, NormFinder calculates intra- and inter-group variation as well as creates a ranking of the most stably expressed genes (Andersen et al., 2004). There is another well-known statistical tool defining the variability of RGs called BestKeeper, which creates the descriptive statistics with several measures and leaves the choice of selection parameters to the experimenter (Pfaffl et al., 2004). Literature regarding the selection of RGs in chicken tissues is scarce. Existing reports are related to infectious diseases (Kuchipudi et al., 2012; Yang et al., 2013; Yin et al., 2011; Yue et al., 2010), immunological research (Borowska et al., 2016; De Boever et al., 2008; Mitra et al., 2016) or different tissues (Bagés et al., 2015; Nascimento et al., 2015), while there are no reports associated with toxicological studies. In this type of research the selection of a proper internal control gene is extremely important due to the significant impact of
Corresponding author. E-mail address: [email protected] (D. Katarzyńska-Banasik).
http://dx.doi.org/10.1016/j.etap.2017.09.011 Received 5 June 2017; Received in revised form 17 September 2017; Accepted 18 September 2017 Available online 20 September 2017 1382-6689/ © 2017 Published by Elsevier B.V.
Environmental Toxicology and Pharmacology 56 (2017) 186–190
D. Katarzyńska-Banasik et al.
Table 1 The summary of TaqMan Gene Expression Assays characteristic. Gene symbol
Assay ID
Context sequence
Amplicon size (bp)
SDHA RPL13 TBP HMBS VIM GAPDH HPRT1 CYP19A1 18S rRNA
Gg03330765_m1 Gg03348054_m1 Gg03366488_m1 Gg03317022_m1 Gg03360311_m1 Gg03346982_m1 Gg03338899_m1 Gg03346001_m1 Eukaryotic 18S rRNA Endogenous Control
GCAGAAGACAATGCAAAGCCATGCT TTATGCCGATCAGGAACGTTTTCAA CAGGAGCAAAAAGCGAGGAACAGTC GCTTTGAGATTGTTGCCATGTCTAC GGAAACTAGAGATGGACAGGTTATT TCGTCAAGCTTGTTTCCTGGTATGA GATATCGGCCAGACTTTGTTGGATT ATTGAAACTGTTATGGGTGACAGAG
103 66 82 87 85 107 89 76 187
SDHA – Succinate dehydrogenase complex flavoprotein subunit A, RPL13 – 60S ribosomal protein L13, TBP – TATA-binding protein, HMBS – Hydroxymethylbilane synthase, VIM –vimentin, GAPDH – Glyceraldehyde 3-phosphate dehydrogenase, HPRT1– Hypoxanthine-guanine phosphoribosyltransferase, CYP19A1 – Cytochrome P450, family 19, subfamily a, polypeptide 1, 18S rRNA – 18S ribosomal RNA.
Transcription Kit (Thermo Scientific) following manufacturer's recommendations. First strand cDNA was synthesized using random primers for 5 min at 25 °C followed by 60 min at 42 °C; reaction was terminated by heating at 70 °C for 5 min. Quantitative real time PCR (qPCR) was performed on 96-well plates in a StepOne Plus thermocycler (Applied Biosystems, Foster City, CA, USA). Each reaction was performed in duplicate in a total volume of 10 μl. One ul of 10-fold diluted cDNA was added to 9 μl of reaction mix containing 5 μl TaqMan Gene Expression Master Mix (Applied Biosystems), 3.5 μl nuclease free water, and 0.5 μl TaqMan Gene Expression Assay (Applied Biosystems). Thermal cycling conditions were as follows: 2 min at 50 °C, 10 min at 95 °C and 40 cycles of denaturation for 15 s at 95 °C, and 1 min of annealing at 60 °C. Standard curves from 10-fold serial dilutions of pooled cDNA were generated to determine amplification efficiency for each tested gene. Amplification efficiencies were as follows: HPRT – 97%, HMBS – 96%, VIM – 99%, 18S rRNA – 95%, GAPDH – 100%, SDHA – 95%, RPL13–102%, TBP – 99%. No template controls with nuclease free water were prepared in each plate in order to check reagent contamination. TaqMan Gene Expression Assays containing primers and probe were chosen from predesigned assays offered by Applied Biosystems (Table 1). Designed probes spanned exon–exon junctions in order to avoid amplification of genomic DNA. Moreover, we used minus-reverse transcriptase controls to assesses the amount of DNA contamination for each gene. 2% agarose gel electrophoresis was prepared in order to verify the amplicons size of each tested gene and it revealed that all gene expression assays amplified a single PCR product with the expected size (Fig. 1). Statistical analysis was carried out using three available software programs: geNorm, NormFinder and BestKeeper. GeNorm calculates the gene expression stability value (M) for the reference gene as the average pairwise variation V for that gene with all other tested reference genes. It creates a ranking of stable RGs by stepwise elimination of genes with the highest M value, determining the two best RGs. GeNorm also indicates the minimum number of genes that should be used for normalization (Vandesompele et al., 2002). This algorithm is sensitive to co-regulation so selected genes must stem from different biological pathways (Andersen et al., 2004). The input data for geNorm were the quantities obtained through transformation of Ct values according to the delta Ct method (Q = E(minCt − sampleCt)). NormFinder is a model based tool, which calculates the M of RGs taking into account intra and inter-group variations. It also suggests the two most stable genes and provides their combined stability value. This model is considered less susceptible to co-regulation (Andersen et al., 2004). Analysis in NormFinder was made on Ct values converted to linear scale according to formula E−ΔCt and then log-transformed by the program itself. BestKeeper creates descriptive statistics with multiple factors, i.e. the geometric mean (GM), arithmetic mean (AM), minimal (Min) and maximal (Max) values, standard deviation (SD), and coefficient of
experimental factors (Arukwe, 2006). The increasing application of silver nanoparticles (AgNPs) in the contemporary world creates the necessity to verify their safety for humans and the environment. AgNPs are now considered as new additives in poultry nutrition and used as disinfectants and odorant limiting agents (Farzinpour and Karashi, 2013; Pineda et al., 2012). In females of domestic birds the number of eggs produced, reproductive rates and breeding results depend on the number of oocytes produced in the ovary, and any anomalies in this area are highly undesirable by breeders. Therefore, the aim of this study was to select the best RGs using normalization of gene expression in the ovaries of hens treated with AgNPs. Eight candidate RGs were analyzed: hypoxanthine-guanine phosphoribosyltransferase (HPRT), hydroxymethylbilane synthase (HMBS), vimentin (VIM), succinate dehydrogenase complex flavoprotein subunit A (SDHA), TATA-binding protein (TBP), 60S ribosomal protein L13 (RPL13), GAPDH, and 18S rRNA. Their stability was assessed using geNorm, NormFinder and BestKeeper software tools.
2. Materials and methods All procedures were performed with the permission of the Local Animal Ethics Committee in Krakow (no. 9/2015). Hy-Line Brown chickens at the age of 25 weeks were kept in individual cages with free access to water and feed, on a 14L:10D lighting schedule. Chickens received per os 1 ml/kg b.w. of colloidal AgNPs in two sizes: 13 nm and 50 nm and in concentrations of 10 ppm or 100 ppm, respectively. The control group was treated with the reference solution in which AgNPs were suspended. After 14-day administration of AgNPs, the chickens were decapitated and prehierachical and hierarchical ovarian follicles were obtained. Granulosa and theca layers were separated from hierarchical ovarian follicles and stored separately. Tissues were snapfrozen in liquid nitrogen immediately after isolation (c.a ∼2–3 min) and stored at −80 °C up to 3 months until RNA extraction from all tissues. All collected samples of ovarian tissues were analyzed together to select the best reference gene for the whole ovary; i.e. from each treatment group (13 nm, 50 nm, control group) we randomly chose 4 samples from prehierachical follicles, 5 granulosa layers from F2 follicles and 5 theca layers from F2 follicles (i.e. 14 biological replicates in each treatment group). It means, that 42 (i.e. 14 replicates ×3 groups) biological repetitions were used in the whole reference gene selection analysis. Total RNA was isolated using silica membrane technology with commercial kit Extractme Total RNA according to the manufacturer's instructions (DNA-Gdańsk, Poland). RNA quality and quantity were measured using NanoDrop Lite (Thermo Scientific, Madison, USA) and checked on 1% agarose gel electrophoresis. The A260/A280 ratios were quantified to assess the quality of the extracted RNA, for all analyzed samples were between 1.96 and 2. Two μg of total RNA was reverse transcribed using a RevertAid RT 187
Environmental Toxicology and Pharmacology 56 (2017) 186–190
D. Katarzyńska-Banasik et al.
Fig. 1. Agarose gel (2%) showing amplification of a specific PCR product of the expected size for the tested genes.
variance (CV), and determines the BestKeeper index (BI). Finally BI is correlated with each RG providing Pearson coefficient correlation (r), coefficient of determination (r2) and p-value. There is no ranking of RGs in BestKeeper and it requires row Ct values to perform the analysis (Pfaffl et al., 2004). CYP19A1 gene, which is one of the steroidogenic genes, was selected to evaluate the reliability of tested reference genes. Amplification efficiency of this gene was 101%. Normalized relative quantity (NRQ) of CYP19A1 gene was calculated using Pfaffl (2001) method and Hellemans et al. (2007) extended model.
0.15
0.1
0.05
3. Results 0
3.1. geNorm Fig. 3. Pairwise variation (Vn/Vn + 1) generated by geNorm for determination of optimal number of reference genes (n = 42).
SDHA and TBP are the most stable reference genes according to geNorm, yielding a combined stability value of M = 0.49 (Fig. 2). Besides geNorm determines the optimal number of RGs by calculating the pairwise variation Vn/Vn+1 between two sequential normalization factors NFn and NFn+1 that contain an increasing number of RGs. Since Vn/Vn+1 recommended threshold is 1.5, this analysis suggests using at least 3 reference genes (V3/4 < 0.15) (Fig. 3). Consequently SDHA, TBP and GAPDH are suggested to be used for normalization as they exhibit the lowest M values. 3.2. NormFinder According to NormFinder the most stable genes are SDHA (M = 0.076) and RPL13 (M = 0.089) (Fig. 4) and their combined stability value is 0.06. Analysis of intra- and inter-group variations for each group are shown in Fig. 5. The columns indicate the inter-group variation and the error bars represent the average of the intra-group variances. This graph clearly demonstrates that the best genes are those with a minimal inter-group variation and the smallest possible error bars.
Fig. 4. Reference genes stability values evaluated by NormFinder (n = 42).
3.3. BestKeeper Descriptive statistics are shown in Table 2. Taking into account SD values HPRT (SD = 0.58), VIM (SD = 0.61), RPL13 (SD = 0.63) and 18S rRNA (SD = 0.75) are the most stable genes, while GAPDH is the worst one (SD = 1.08). However based on correlation coefficient (r), the ranking is as follows: SDHA (r = 0.894), TBP (r = 0.885), GAPDH (r = 0.865), RPL13 (r = 0.854); and 18S rRNA is characterized by the lowest correlation coefficient (r = 0.635). When we combine these two variables, as described in the discussion, SDHA seem to be the best choice, then RPL13 and TBP.
0.90
0.9
0.82
0.8
0.77 0.7
0.68 0.61
0.6
0.54 0.5 0.4 18S
VIM
HPRT
RPL13
HMBS
GAPDH
3.4. Reference gene validation
SDHA TBP
CYP19A1 gene expression was calculated by using all tested genes for normalization (Fig. 6). When RPL13 and SDHA were used for normalization, the expression levels of CYP19A1 gene were equal in
Fig. 2. Average expression stability values of reference genes evaluated by geNorm (n = 42).
188
Environmental Toxicology and Pharmacology 56 (2017) 186–190
D. Katarzyńska-Banasik et al.
greater (it decreased by more than 80% in 50 nm group). The same difference in NRQs was observed when genes selected by geNorm were used for normalization, i.e. SDHA and TBP or SDHA, TBP and GAPDH. Normalization against the most unstable gene 18S rRNA lead to elevated levels of CYP19A1 gene expression in both groups 13 and 50 nm AgNPs.
0.40 0.20 Control
0.00 -0.20 Intr
4. Discussion
-0.40
The necessity to normalize data obtained from real-time PCR toward stable validated reference genes (RGs) is indisputable, especially in toxicological studies. Thus far, we have not found information about the appropriate endogenous control for research conducted on the chicken ovary, nor any toxicological studies in relation to this organ. Olias et al. (2014) evaluated RGs in the ovarian tissue of song birds (such as zebra finches), but there are no reports on the validation of RGs in chicken tissues, where the influence of toxic agents has been investigated. Except for genes selected on the basis of research conducted on zebra finches (Olias et al., 2014), we have included two commonly used internal controls: GAPDH and 18S rRNA. These genes served as individual RGs for normalization of qPCR data in multiple studies in chickens (Adams et al., 2009; Hong et al., 2006; Hrabia et al., 2013; Richards et al., 2006; Zhang et al., 2012). In our study, 18S rRNA was revealed to be the least stable gene by geNorm and NormFinder. In BestKeeper, 18S rRNA had the worst coefficient correlation, while GAPDH was characterized by the highest SD. Analysis conducted separately in control and each experimental group confirmed that the 18S rRNA gene is highly unstable (data not shown). That is in agreement with the previous findings in avian gonads (Olias et al., 2014). Generally, the 18S rRNA gene has some disadvantages and in many cases is not recommended for normalization (Kozera and Rapacz, 2013). Results regarding the most stable expressed genes were similar in BestKeeeper and NormFinder analyses, but entirely different in geNorm. BestKeeper does not specify which genes are the most stable and the experimenter has to decide for himself which parameter is the most meaningful. Although the combination of SD and the correlation coefficient of BI with each RG is predominantly used (Olias et al., 2014; Wood et al., 2008; Normann et al., 2016), in some studies only one of these variables is taken into account: only SD (Wang et al., 2012; Ostrowska et al., 2014) or only r (Maroufi et al., 2010). We decided to combine SD and coefficient correlation. Following the suggestion of Pfaffl et al. (2004) we excluded genes with SD higher than 1, such as GAPDH. Then we decided to rank genes based on SD and coefficient of correlation higher than 0.8. Implementing this method SDHA, RPL13 and TBP were considered as the most stably expressed genes. The same sequence of genes was indicated by NormFinder, which suggested that SDHA and RPL13 together were the best combination of genes for qPCR data normalization. Another pair of genes was selected by geNorm, namely the SDHA and TPB. The reason for these differences are diverse calculation strategies used in these programs. GeNorm uses a pairwise comparison approach and selects two genes with lowest intragroup variation and close enough intergroup variation. Consequently, the selected genes will have similar expression profiles but they will not necessarily be the most stable genes (Andersen et al., 2004). On the other hand, NormFinder selects two candidates with minimal shared inter- and intragroup variation in a so-called modelbased approach (Andersen et al., 2004). The model-based strategy seems to be more precise and trustworthy, so we decided to make use of SDHA and RPL13 as the most stable genes identified by both BestKeeper and NormFinder for further analysis of gene expression. Our choice was confirmed by CYP19A1 gene expression calculations and the analysis showed that normalization against only one of the selected genes SDHA or RPL13 would not generate different expression patterns than normalization against both of them. CYP19A1 gene expression also revealed that normalization using the genes selected by GeNorm (SDHA, TBP) yields other expression results than normalization with respect to
-0.60
Fig. 5. Inter-and intra-group variation evaluated by NormFinder. The graph bars and the error bars represent the inter-group variation and an average of the intra-group variation, respectively (n = 14 in each group).
Table 2 BestKeeper descriptive statistics of the analyzed genes in the chicken ovary. Statistics derived from the 42 biological replicates.
n gmean SD [ ± Ct] CV [%Ct] Bk [r] p-value
HPRT
VIM
RPL13
18S
HMBS
SDHA
TBP
GAPDH
42 26.31 0.58 2.19 0.737 0.001
42 24.49 0.61 2.51 0.667 0.001
42 25.11 0.63 2.53 0.854 0.001
42 13.14 0.75 5.69 0.635 0.001
42 27.00 0.85 3.14 0.796 0.001
42 28.63 0.86 3.01 0.894 0.001
42 28.63 0.97 3.38 0.885 0.001
42 25.03 1.08 4.31 0.865 0.001
SDHA – Succinate dehydrogenase complex flavoprotein subunit A, RPL13 – 60S ribosomal protein L13, TBP – TATA-binding protein, HMBS – Hydroxymethylbilane synthase, VIM – vimentin, GAPDH – Glyceraldehyde 3-phosphate dehydrogenase, HPRT1– Hypoxanthineguanine phosphoribosyltransferase, 18S rRNA – 18S ribosomal RNA, gmean – geometric mean, SD [ ± Ct] – standard deviation of the Ct, CV [%Ct] – the coefficient of variance expressed as a percentage of the Ct level, Bk [r] – Pearson’s coefficient correlation of each gene to the BI-index, p-value – probability value.
Fig. 6. Relative quantification of CYP19A1 gene expression using tested reference genes for normalization under AgNPs treatment in chicken ovary. Each value represents the mean ± SEM from five determinations.
control and 13 nm AgNPs group and decreased in 50 nm AgNPs group by approximately 30%. The same expression patterns were observed when RPL13 and SDHA were used separately. When less stable reference genes TBP, HMBS, GAPDH, HPRT and VIM were used for normalization the expression levels increased in 13 nm AgNPs group and the difference between NRQ in 13 nm group and 50 nm group was 189
Environmental Toxicology and Pharmacology 56 (2017) 186–190
D. Katarzyńska-Banasik et al.
Lekanne Deprez, R.H., Fijnvandraat, A.C., Ruijter, J.M., Moorman, A.F.M., 2002. Sensitivity and accuracy of quantitative real time polymerase chain reaction using SYBR green I depends on cDNA synthesis conditions. Anal. Biochem. 307, 63–69. Maroufi, A., Van Bockstaele, E., De Loose, M., 2010. Validation of reference genes for gene expression analysis in chicory (Cichorium intybus) using quantitive real-time PCR. BMC Mol. Biol. 11, 1–12. Mitra, T., Bilic, I., Hess, M., Liebhart, D., 2016. The 60S ribosomal protein L13 is the most preferable reference gene to investigate gene expression in selected organs from turkeys and chickens, in context of different infection models. Vet. Res. 47, 105. Nascimento, C.S., Barbosa, L.T., Brito, C., Fernandes, R.P.M., Mann, R.S., Pinto, A.P., Oliveira, H.C., Dodson, M.V., Guimarães, S.E.F., Duarte, M.S., 2015. Identification of suitable reference genes for real time quantitative polymerase chain reaction assays on Pectoralis major muscle in chicken (Gallus gallus). PLoS One 10 (5), e0127935. Nicot, N., Hausman, J.F., Hoffmann, L., Evers, D., 2005. Housekeeping gene selection for real-time RT-PCR normalization in potato during biotic and abiotic stress. J. Exp. Bot. 56, 2907–2914. Normann, K.R., Øystese, K.A., Berg, J.P., Lekva, T., Berg-Johnsen, J., Bollerslev, J., Olarescu, N.C., 2016. Selection and validation of reliable reference genes for RTqPCR analysis in a large cohort of pituitary adenomas. Mol. Cell. Endocrinol. 437, 183–189. Ohl, F., Jung, M., Xu, C., Stephan, C., Rabien, A., Burkhardt, M., Nitsche, A., Kristiansen, G., Loening, S.A., Radonic, A., Jung, K., 2005. Gene expression studies in prostate cancer tissue: which reference gene should be selected for normalization? J. Mol. Med. 83, 1014–1024. Olias, P., Adam, I., Meyer, A., Scharff, C., Gruber, A.D., 2014. Reference genes for quantitative gene expression studies in multiple avian species. PLoS One 9, e99678. Ostrowska, M., Żelazowska, B., Słoniewski, K., Kowalski, Z.M., Zwierzchowski, L., 2014. Technical note: selecting the best references in gene expression experiments in liver of cows receiving glucogenic supplements during the transition period. J. Dairy Sci. 97, 911–916. Pfaffl, M.W., Tichopad, A., Prgomet, C., Neuvians, T.P., 2004. Determination of stable housekeeping genes, differentially regulated target genes and sample integrity: BestKeeper–excel-based tool using pair-wise correlations. Biotechnol. Lett. 26, 509–515. Pfaffl, M.W., 2001. A new mathematical model for relative quantification in real-time RTPCR. Nucleic Acids Res. 29 (9), e45. Pineda, L., Chwalibog, A., Sawosz, E., Lauridsen, C., Engberg, R., Elnif, J., Hotowy, A., Sawosz, F., Gao, Y., Ali, A., Moghaddam, H.S., 2012. Effects of silver nanoparticles on growth performance, metabolism and microbial profile of broiler chickens. Arch. Anim. Nutr. 66, 416–429. Radonic, A., Thulke, S., Mackay, I.M., Landt, O., Siegert, W., Nitsche, A., 2004. Guideline to reference gene selection for quantitative real-time PCR. Biochem Biophys Res. Commun. 313, 856–862. Richards, M.P., Poch, S.M., McMurtry, J.P., 2006. Characterization of turkey and chicken ghrelin genes, and regulation of ghrelin and ghrelin receptor mRNA levels in broiler chickens. Gen. Comp. Endocrinol. 145, 298–310. Sanders, R., Mason, D., Foy, C., Huggett, J., 2014. Considerations for accurate gene expression measurement by reverse transcription quantitative PCR when analysing clinical samples. Anal. Bioanal. Chem. 406, 6471–6483. Selvey, S., Thompson, E.W., Matthaei, K., Le, A.R.A., Irving, M.G., Griffiths, L.R., 2001. Beta-actin—an unsuitable internal control for RT-PCR. Mol. Cell. Probes 15, 307–311. Suzuki, T., Higgins, P.J., Crawford, D.R., 2000. Control selection for RNA quantitation. Biotechniques 29, 332–337. Thellin, O., Zorzi, W., Lakaye, B., De Borman, B., Coumans, B., Hennen, G., Grisar, T., Igout, A., Heinen, E., 1999. Housekeeping genes as internal standards: use and limits. J. Biotechnol. 75, 291–295. Thellin, O., ElMoualij, B., Heinen, E., Zorzi, W., 2009. A decade of improvements in quantification of gene expression and internal standard selection. Biotechnol. Adv. 27, 323–333. Toegel, S., Huang, W., Piana, C., Unger, F.M., Wirth, M., Goldring, M.B., Gabor, F., Viernstein, H., 2007. Selection of reliable reference genes for qPCR studies on chondroprotective action. BMC Mol. Biol. 8, 13. Vandesompele, J., De Preter, K., Pattyn, F., Poppe, B., Van Roy, N., De Paepe, A., Speleman, F., 2002. Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol. 3 (research0034):1. Wang, Q., Ishikawa, T., Michiue, T., Zhu, B.L., Guan, D.W., Maeda, H., 2012. Stability of endogenous reference genes in postmortem human brains for normalization of quantitative real-time PCR data: comprehensive evaluation using geNorm NormFinder, and BestKeeper. Int. J. Leg. Med. 126, 943–952. Wong, M.L., Medrano, J.F., 2005. Real time PCR for mRNA quantification. Biotechniques 39, 75–85. Wood, S.H., Clements, D.N., McEwan, N.A., Nuttall, T., Carter, S.D., 2008. Reference genes for canine skin when using quantitative real-time PCR. Vet. Immunol. Immunopathol. 126, 392–395. Yang, F., Lei, X., Rodriguez-Palacios, A., Tang, C., Yue, H., 2013. Selection of reference genes for quantitative real-time PCR analysis in chicken embryo fibroblasts infected with avian leukosis virus subgroup. J. BMC Res. Notes 6, 402. Yin, R., Liu, X., Liu, C., Ding, Z., Zhang, X., Tian, F., Liu, W., Yu, J., Li, L., Hrabé de Angelis, M., Stoeger, T., 2011. Systematic selection of housekeeping genes for gene expression normalization in chicken embryo fibroblasts infected with Newcastle disease virus. Biochem. Biophys. Res. Comm. 413, 537–540. Yue, H., Lei, X.W., Yang, F.L., Li, M.Y., Tang, C., 2010. Reference gene selection for normalization of PCR analysis in chicken embryo fibroblast infected with H5N1 AIV. Virol Sin. 25, 425–431. Zhang, W., Sargis, R.M., Volden, P.A., Carmean, C.M., Sun, X.J., Brady, M.J., 2012. PCB 126 and other dioxin-like PCBs specifically suppress hepatic PEPCK expression via the aryl hydrocarbon receptor. PLoS One 7 (5), e37103.
the genes recommended by NormFinder. It is not surprising because SDHA and TBP showed the same nonvanishing intergroup variation. 5. Conclusions In summary, the stability of RGs changes under different experimental conditions. This is the first report undertaken to select RGs for silver nanoparticle toxicity studies in the chicken ovary. In this study, we have determined that 18S rRNA is not stable in the chicken ovarian tissues and we do not recommend utilizing it in such studies. From the eight analyzed genes, we have chosen SDHA and RPL13 for further qPCR analysis. Conflict of interest The authors declare that there is no conflict of interest regarding the publication of this paper. Acknowledgments This study was supported by the National Science Centre, Poland [grant number UMO-2014/15/N/NZ9/01435]. The authors thank Professor Anna Hrabia and Dr Maria Mika for technical help during the experiment, Ms. Kinga Kowalik for assisting in RNA isolation and Mrs Balbina Sobesto for animal care. References Adams, S.C., Xing, Z., Li, J., Cardona, C.J., 2009. Immune-related gene expression in response to H11N9 low pathogenic avian influenza virus infection in chicken and Pekin duck peripheral blood mononuclear cells. Mol. Immunol. 46, 1744–1749. Andersen, C.L., Jensen, J.L., Ørntoft, T.F., 2004. 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 Res. 64, 5245–5250. Arukwe, A., 2006. Toxicological housekeeping genes: do they really keep the house? Environ. Sci. Technol. 40, 7944–7949. Bagés, S., Estany, J., Tor, M., Pena, R.N., 2015. Investigating reference genes for quantitative real-time PCR analysis across four chicken tissues. Gene 561, 82–87. Borowska, D., Rothwell, L., Bailey, R.A., Watson, K., Kaiser, P., 2016. Identification of stable reference genes for quantitative PCR in cells derived from chicken lymphoid organs. Vet. Immunol. Immunopathol. 170, 20–24. Bustin, S.A., Benes, V., Nolan, T., Pfaffl, M.W., 2005. Quantitative real-time RT-PCR–a perspective. J. Mol. Endocrinol. 34, 597–601. Bustin, S.A., Benes, V., Garson, J.A., Hellemans, J., Huggett, J., Kubista, M., Mueller, R., Nolan, T., Pfaffl, M.W., Shipley, G.L., Vandesompele, J., Wittwer, C.T., 2009. The MIQE guidelines: minimum information for publication of quantitative real-time PCR experiments. Clin. Chem. 55, 611–622. Bustin, S.A., 2002. Quantification of mRNA using real time reverse transcription PCR (RTPCR): trends and problems. J. Mol. Endocrinol. 29, 23–39. De Boever, S., Vangestel, C., De Backer, P., Croubels, S., Sys, S.U., 2008. Identification and validation of housekeeping genes as internal control for gene expression in an intravenous LPS inflammation model in chickens. Vet. Immunol. Immunopathol. 122, 312–317. Derveaux, S., Vandesompele, J., Hellemans, J., 2010. How to do successful gene expression analysis using real-time PCR. Methods 50, 227–230. Farzinpour, A., Karashi, N., 2013. The effects of silver nanoparticles on egg quality traits in laying Japanese quail. Appl Nanosci. 3, 95–99. Glare, E.M., Divjak, M., Bailey, M.J., Walters, E.H., 2002. Beta-actin and GAPDH housekeeping gene expression in asthmatic airways is variable and not suitable for normalising mRNA levels. Thorax 57, 765–770. Hellemans, J., Mortier, G., De Paepe, A., Spelman, F., Vandesompele, J., 2007. qBase relative quantification framework and software for management and automated analysis of real-time quantitative PCR data. Genome Biol. 8 (2), R19. Hong, Y.H., Lillehoj, H.S., Lillehoj, E.P., Lee, S.H., 2006. Changes in immune-related gene expression and intestinal lymphocyte subpopulations following Eimeria maxima infection of chickens. Vet. Immunol. Immunopathol. 114, 259–272. Hrabia, A., Leśniak, A., Sechman, A., 2013. In vitro effects of TCDD, PCB126 and PCB153 on estrogen receptors, caspases and metalloproteinase-2 mRNA expression in the chicken shell gland. Folia Biol. 61, 3–4. Huggett, Y., Dheda, K., Bustin, S., Zumala, A., 2005. Real-time RT-PCR normalization; strategies and considerations. Genes Immun. 6, 279–284. Kozera, B., Rapacz, M., 2013. Reference genes in real-time PCR. J. Appl. Genet. 54, 391–406. Kuchipudi, S.V., Tellabati, M., Nelli, R.K., White, G.A., Perez, B.B., Sebastian, S., Slomka, M.J., Brookes, S.M., Brown, I.H., Dunham, S.P., 2012. 18S rRNA is a reliable normalisation gene for real time PCR based on influenza virus infected cells. Virol. J. 9, 230.
190