Fish & Shellfish Immunology 36 (2014) 374e382
Contents lists available at ScienceDirect
Fish & Shellfish Immunology journal homepage: www.elsevier.com/locate/fsi
Full length article
Determination of reference microRNAs for relative quantification in grass carp (Ctenopharyngodon idella) Xiao-Yan Xu a, Yu-Bang Shen a, Jian-Jun Fu a, Li-Qun Lu c, Jia-Le Li a, b, * a
Key Laboratory of Exploration and Utilization of Aquatic Genetic Resources, Shanghai Ocean University, Ministry of Education, Shanghai 201306, PR China E-Institute of Shanghai Universities, Shanghai Ocean University, 999 Huchenghuan Road, 201306 Shanghai, PR China c National Pathogen Collection Center for Aquatic Animals, College of Fisheries and Life Science, Shanghai Ocean University, 999 Huchenghuan Road, 201306 Shanghai, PR China b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 14 August 2013 Received in revised form 30 November 2013 Accepted 16 December 2013 Available online 22 December 2013
Relative quantification is the strategy of choice for processing real-time quantitative reverse transcription polymerase chain reaction (RT-qPCR) data in microRNA (miRNA) expression studies. Normalization of relative quantification data is performed by comparison to reference genes. In teleost species, such as grass carp (Ctenopharyngodon idella), the determination of reference miRNAs and the optimal numbers of these that should be used has not been widely studied. In the present study, the stability of seven miRNAs (miR-126-3p, miR-101a, miR-451, miR-22a, miR-146, miR-142a-5p and miR-192) was investigated by RT-qPCR in different tissues and in different development stages of grass carp. Stability values were calculated with geNorm, NormFinder, BestKeeper and Delta CT algorithms. The results showed that tissue type is an important variability factor for miRNA expression stability. All seven miRNAs had good stability values and, therefore, could be used as reference miRNAs. When all tissues and developmental stages were considered, miR-101a was the most stable miRNA. When each tissue type was considered separately, the most stable miRNAs were 126-3p in blood and liver, 101a in the gills, 192 in the kidney, 451 in the intestine and 22a in the brain, head kidney, spleen, heart, muscle, skin and fin. 126-3p was the most stable reference miRNA gene during developmental stages 1e5, while 22a was the most stable during developmental stages 6e18. Overall, this study provides valuable information about the reference miRNAs that can be used to perform appropriate normalizations when undertaking relative quantification in RT-qPCR studies of grass carp. Ó 2013 Elsevier Ltd. All rights reserved.
Keywords: Internal control Real-time PCR Grass carp MicroRNA
1. Introduction MicroRNAs (miRNAs) are w22 nucleotide RNAs that regulate gene expression through translational repression and/or transcript cleavage [1] in organisms as diverse as viruses, unicellular algae, plants, worms, flies, fish and mammals [2,3]. miRNAs are essential for vertebrate development and are likely to be involved in differentiation and/or maintenance of tissue and cell identity. The current set of miRNAs is predicted to regulate several thousands of target mRNAs, which may include up to 30% of all protein-coding genes [4]. Therefore, there is strong interest in analysing the roles
Abbreviations: EF1a, elongation factor 1a; GAPDH, glycer-aldehyde-3-phosphate dehydrogenase; miRNA, microRNA; RT-qPCR, real-time quantitative reverse transcription polymerase chain reaction. * Corresponding author. Key Laboratory of Exploration and Utilization of Aquatic Genetic Resources, Shanghai Ocean University, Ministry of Education, Shanghai 201306, PR China. Tel.: þ86 021 61900401; fax: þ86 021 61900405. E-mail address: [email protected] (J.-L. Li). 1050-4648/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.fsi.2013.12.007
of miRNAs in development, disease and other cellular processes. miRNAs negatively regulate gene expression through two major mechanisms: translational repression and mRNA cleavage, which depend on the extent of complementarity between the miRNA and its mRNA target [5]. Since miRNAs seem to regulate gene expression by a mechanism of ‘fine-tuning’, the study of the participation of miRNAs and their targets in specific physiological or pathological experimental situations depends heavily on a reliable and accurate technique for measuring miRNA expression levels [6]. Real-time quantitative reverse transcription polymerase chain reaction (RT-qPCR) is an accurate and sensitive technique for performing gene expression analysis. The use of reference genes as internal controls for the measurement of target gene expression variation is the preferred method for normalizing RT-qPCR data because this approach captures all non-biological variation [7]. The selection of the reference gene(s) to use is not trivial and previous studies have demonstrated that a single universal reference gene is unlikely to exist and perform well for all tissue types or for all physiological, pathological and experimental situations [8,9].
X.-Y. Xu et al. / Fish & Shellfish Immunology 36 (2014) 374e382
375
Originally, normalization strategies were performed using only one reference gene; however, this idea has now evolved to include different normalization approximations, such as using the global mean normalization method or the robust multiple reference gene normalization approach where more than one reference gene is used [10]. To date, several genes, including 18S rRNA, b-actin, elongation factor 1a (EF1a) and glycer-aldehyde-3-phosphate dehydrogenase (GAPDH), have been used widely as candidate reference genes for gene expression studies in grass carp (Ctenopharyngodon idella) [11,12]. In miRNA expression studies, the most common reference genes used are ribosomal RNAs, such as 18S rRNA [13], and small nuclear RNAs like U6 snRNA [14]. However, the use of miRNAs as reference genes is not widely used at present, although it is very important that the references used have the same nature as the study subjects. The reference genes used should have the same length as the molecules of interest in order to assure the same efficiency during RNA isolation and reverse transcription [15]. In this sense, only a few studies have explored the stability of some miRNAs in human, rat and porcine tissues [6,10,16,17]. To date, no one report has deeply analysed the expression stability of miRNAs to be used as references in teleost studies. The aim of this present work was to identify suitable reference gene(s) and minimize the risk of co-regulation artifacts in grass carp. The expression of seven candidate reference miRNA genes (101a, 126-3p, 146, 192, 22a, 451 and 142a-5p) were examined in 18 developmental stages (unfertilized eggs; 0 h post-fertilization; embryos at the 16-cell stage; morula stage; gastrula stage; eye sac-appearance stage; caudal fin-appearance stage; muscular effect stage; heart beating stage; mental stages; and at 1, 2, 3, 4, 5, 6, 7 and 10 days post-hatching.) and 12 tissues (blood, brain, muscle, trunk kidney, liver, head kidney, skin, spleen, heart, gill, intestine and fin) from control fish and fish challenged with bacteria. Furthermore, the consistency of the best-scoring reference gene was tested by four statistical approaches (geNorm, BestKeeper, NormFinder and Delta CT algorithms).
total biomass. Embryos and fries were also obtained from the Wujiang National Farm of Chinese Four Family Carps, and reared in a hatching trough with constant pool water flow at 21 1 C. On day 3 post-hatching, the fry could swim steadily and were fed with freshwater rotifers captured from the pool. Twelve tissue samples were collected from three grass carp (blood, brain, muscle, trunk kidney, liver, head kidney, skin, spleen, heart, gill, intestine and fin), while embryos and early larvae were sampled at 18 different developmental stages (10 specimens from each stage). These stages were: unfertilized eggs; 0 h postfertilization; embryos at the 16-cell stage; morula stage; gastrula stage; eye sac-appearance stage; caudal fin-appearance stage; muscular effect stage; heart beating stage; mental stages; and at 1, 2, 3, 4, 5, 6, 7 and 10 days post-hatching. For the bacterial challenge, 12 fish were intrapleurally injected with formalin-killed Aeromonas hydrophila S2 (obtained from the Aquatic Pathogen Collection Centre of the Ministry of Agriculture, China) at a dose of 7.0 106 cells suspended in 100 ml PBS per fish; 12 control fish were similarly injected with 100 ml sterile PBS per fish. Three fish were sampled at 4 h, 1 day, 3 days and 7 days postinjection, respectively. Blood, brain, muscle, trunk kidney, liver, head kidney, skin, spleen, heart, gill, intestine and fin were collected from each fish. All samples were immediately snap-frozen in liquid nitrogen and stored at 80 C until use. Two groups were maintained in two aquariums and intraperitoneally injected with A. hydrophila AH10 (Aquatic Pathogen Collection Centre of Ministry of Agriculture, China) at a dose of 7.0 106 cells suspended in 100 ml PBS per fish. All the fish were observed every 4 h for any mortality and collecting samples until the termination of the experiment at 240 h post-challenge. Grass carp that died in the first 72 h post-challenge were classified as susceptible group (SG), while the animals that survived over 240 h post-challenge were considered as resistant group (RG).
2. Materials and methods
The candidate reference genes selected for evaluation included miR-148, miR-192, miR-451, miR-126-3p, miR-101a, miR-142a-5p, miR-146b and miR-22a. These miRNA candidate genes were selected due to their relative high quantities in miRNA profiles of grass carp that were generated during Illumina deep sequencing (data not shown). 18S rRNA and b-actin mRNAs were selected to be evaluated as reference genes due to their wide use in the literature. This selection process was performed after considering stability values reported in previous studies [11,12].
2.1. Sample collection Grass carp with an average weight of 50 g were cultured individually in Wujiang National Farm of Chinese Four Family Carps (Jiangsu Province, China). Animals were raised at 28 C in 400-L aerated tanks for one week before the experiment and fed twice daily (in the morning and late in the afternoon) at a ratio of 5% of
2.2. Selection of candidate reference miRNAs
Table 1 Sequence information for 14 selected candidate reference genes. Gene symbol
Primer sequence (50 e30 )
Amplification efficiency
miR-148 miR-192 miR-451 miR-let-7a miR-126-3p miR-101a miR-142a-5p miR-146b miR-22a miR-217 miR-142a-3p miR-23a B-actin (b-actin)
TCAGTGCATTACAGAACTTTGT ATGACCTATGAATTGACAGCC AAACCGTTACCATTACTGAGTT TGAGGTAGTAGGTTGTATAGTT CTCGTACCGTGAGTAATAATGC TACAGTACTGTGATAACTGAAG CATAAAGTAGACAGCACTACT TGAGAACTGAATTCCAAGGGTG AAGCTGCCAGCTGAAGAACTGT TACTGCATCAGGAACTGATTGG CGTGTAGTGTTTCCTACTTTATGGA ATCACATTGCCAGGGATTTCCA Forward: CCTTCTTGGGTATGGAATCTTG Reverse: AGAGTATTTACGCTCAGGTGGG Forward: GGACACGGAAAGGATTGACAG Reverse: CGGAGTCTCGTTCGTTATCGG
87.5 95.4 95.7 96.0 96.3 96.6 99.9 104.8 105.0 97.8 96.0 96.5 95.8
18S rRNA (18S ribosomal RNA)
98.2
376
X.-Y. Xu et al. / Fish & Shellfish Immunology 36 (2014) 374e382
Table 2 Ranking of candidate reference genes according to stability values produced by geNorm, NormFinder, BestKeeper and Delta CT, and the recommended comprehensive ranking calculated by RefFinder. A. Rank order of each miRNA gene in the different tissues of control fish. B. The geNorm stability value of each miRNA gene in the different tissues of control fish. A Ranking order (betteregoodeaverage) Method
1
2
3
4
5
6
7
8
9
Delta CT BestKeeper Normfinder geNorm Recommended comprehensive ranking
101a 22a 101a 146 j 142a-5p 101a
142a-5p 101a 142a-5p
146 18s 146 101a 146
22a 126-3p 22a 126-3p 22a
126-3p 146 126-3p 22a 126-3p
18s 142a-5p 18s 18s 18s
B-actin B-actin B-actin B-actin B-actin
451 451 451 451 451
192 192 192 192 192
142a-5p
B Gene name
146 j 142a-5p
101a
126-3p
22a
18s
B-actin
451
192
geNorm Stability value
0.87
1.222
1.503
1.639
1.761
2.108
2.358
2.835
to calculate the RT-qPCR efficiency. All standard curves were generated using 10-fold serial dilutions from a pool of cDNA from all the samples (n ¼ 30). Reactions were incubated in a 96-well plate at 95 C for 10 min, followed by 40 cycles of 95 C for 3 s and 60 C for 30 s on a CFX96 Real-Time PCR System (Bio-Rad, CA, USA). Details of all primers used are listed in Table 1. All measurements were performed in duplicate. Minus RT controls, minus poly (A) polymerase controls and no template controls were each included. Moreover, melting curve analysis was performed in each assay in order to detect non-specific amplifications.
2.3. RNA isolation and cDNA synthesis Total RNA was isolated using TRIzol reagent following the manufacturer’s recommendations (Invitrogen, Carlsbad, USA) and stored at 80 C after incubation with RNase-free gDNA Eraser (TaKaRa, Japan). The RNA concentration and purity was determined by measuring the absorbance at 260 nm and 280 nm in a Nanodrop 2000 spectrophotometer (Nanodrop Technologies, Wilmington, DE). Reversely transcribed using One Step PrimeScript miRNA cDNA Synthesis Kit (TaKaRa). This kit adds poly (A) to the 30 end of miRNA and start to reversely transcribe. The reverse transcription was led by a kind of special oligo-dT ligated with known sequence at its 50 end.
2.5. Stability expression analysis The stability of each candidate miRNA was tested using geNorm v.3.5 [9], NormFinder [18] BestKeeper [19] and the comparative delta-Ct method [20] algorithm. The geNorm algorithm calculates the gene expression stability (M) value for each candidate reference gene based on average pairwise variation between all studied genes. NormFinder is based on an analysis of variance (ANOVA) mathematical model and this estimates intra- and inter-group
2.4. RT-qPCR RT-qPCR was performed in a final volume of 20 mL and each reaction included 10 mL of SYBR Premix Ex Taq Ⅱ (2x) (Takara), 0.4 mM of Uni-miR qPCR Primer, 0.4 mM of Forward Primer and 2 mL of a 1:5 dilution of cDNA. Standard curves were generated in order
Table 3 Ranking of candidate reference genes according to stability values in different tissues of fish after bacterial challenge. The stability measurements were produced by geNorm, NormFinder, BestKeeper and Delta CT, and the recommended comprehensive ranking was calculated by RefFinder. A. Rank order of each miRNA gene in the different tissue. B. Rank order and geNorm stability value of each miRNA gene in spleen and kidney between SG and RG sample. A Ranking order (betteregoodeaverage) Tissue
1
2
3
4
5
6
7
8
9
Blood Brain Kidney Head kidney Liver Spleen Gill Heart Muscle Skin Intestine Fin
126-3p 22a 192 22a 126-3p 22a 101a 22a 22a 22a 451 22a
101a 126-3p 126-3p 142a-5p 192 101a 22a 101a 142a-5p 126-3p 142a-5p 451
451 146 22a 101a 451 126-3p 126-3p 126-3p 451 101a 22a 126-3p
22a 451 451 126-3p 22a 192 142a-5p 142a-5p 101a 146 146 18s
146 101a 146 146 101a 146 451 451 126-3p 451 192 101a
142a-5p 142a-5p 101a 18s 146 142a-5p 146 18s 18s 142a-5p 101a 142a-5p
18s 192 142a-5p 451 142a-5p 451 192 192 146 18s 126-3p 146
192 18s 18s 192 18s 18s 18s 146 B-actin 192 18s 192
B-actin B-actin B-actin B-actin B-actin B-actin B-actin B-actin 192 B-actin B-actin B-actin
B Ranking order (betteregoodeaverage) Gene name KidneygeNorm stability value Gene name SpleengeNorm stability value
126-3p 0.509 22a 0.758
192 0.535 192 0.734
146 0.572 101a 0.826
22a 0.572 126-3p 0.740
451 0.665 146 1.121
101a 0.715 142a-5p 1.166
142a-5p 1.269 451 2.300
X.-Y. Xu et al. / Fish & Shellfish Immunology 36 (2014) 374e382
377
the lowest biological variance. As no unique criterion exists to study this issue, we applied the four most frequently used programs: BestKeeper, geNorm, NormFinder and the comparative delta-Ct method. Based on the rankings from each of these programs, RefFinder calculated an overall final ranking and the results are shown in Tables 2e4. Across the different tissues examined in the miRNA studies, miR-101a was the most stably expressed gene (Table 2). The expression stability of each miRNA gene in the different RNA samples tested at various time points post-bacterial infection was evaluated by RefFinder. The most stable miRNAs were 126-3p in the blood and liver, 101a in the gill, 192 in the kidney, 451 in the intestine and 22a in the brain, head kidney, spleen, heart, muscle, skin and fin (Table 3A). Between SG and RG samples, we found same result, 192 and 126 in the kidney and 22a in the spleen was the most stable miRNAs (Table 3B). During the different developmental stages, the order of gene expression stability was 101a, 22a, 146, 142a-5p, 451, 126-3p, 18s, 192 and, finally, b-actin (Table 4). However, geNorm showed that all the miRNAs had high M values (>1.5), except for 101a and 22a. Gene with the lowest M values (
Contents lists available at ScienceDirect
Fish & Shellfish Immunology journal homepage: www.elsevier.com/locate/fsi
Full length article
Determination of reference microRNAs for relative quantification in grass carp (Ctenopharyngodon idella) Xiao-Yan Xu a, Yu-Bang Shen a, Jian-Jun Fu a, Li-Qun Lu c, Jia-Le Li a, b, * a
Key Laboratory of Exploration and Utilization of Aquatic Genetic Resources, Shanghai Ocean University, Ministry of Education, Shanghai 201306, PR China E-Institute of Shanghai Universities, Shanghai Ocean University, 999 Huchenghuan Road, 201306 Shanghai, PR China c National Pathogen Collection Center for Aquatic Animals, College of Fisheries and Life Science, Shanghai Ocean University, 999 Huchenghuan Road, 201306 Shanghai, PR China b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 14 August 2013 Received in revised form 30 November 2013 Accepted 16 December 2013 Available online 22 December 2013
Relative quantification is the strategy of choice for processing real-time quantitative reverse transcription polymerase chain reaction (RT-qPCR) data in microRNA (miRNA) expression studies. Normalization of relative quantification data is performed by comparison to reference genes. In teleost species, such as grass carp (Ctenopharyngodon idella), the determination of reference miRNAs and the optimal numbers of these that should be used has not been widely studied. In the present study, the stability of seven miRNAs (miR-126-3p, miR-101a, miR-451, miR-22a, miR-146, miR-142a-5p and miR-192) was investigated by RT-qPCR in different tissues and in different development stages of grass carp. Stability values were calculated with geNorm, NormFinder, BestKeeper and Delta CT algorithms. The results showed that tissue type is an important variability factor for miRNA expression stability. All seven miRNAs had good stability values and, therefore, could be used as reference miRNAs. When all tissues and developmental stages were considered, miR-101a was the most stable miRNA. When each tissue type was considered separately, the most stable miRNAs were 126-3p in blood and liver, 101a in the gills, 192 in the kidney, 451 in the intestine and 22a in the brain, head kidney, spleen, heart, muscle, skin and fin. 126-3p was the most stable reference miRNA gene during developmental stages 1e5, while 22a was the most stable during developmental stages 6e18. Overall, this study provides valuable information about the reference miRNAs that can be used to perform appropriate normalizations when undertaking relative quantification in RT-qPCR studies of grass carp. Ó 2013 Elsevier Ltd. All rights reserved.
Keywords: Internal control Real-time PCR Grass carp MicroRNA
1. Introduction MicroRNAs (miRNAs) are w22 nucleotide RNAs that regulate gene expression through translational repression and/or transcript cleavage [1] in organisms as diverse as viruses, unicellular algae, plants, worms, flies, fish and mammals [2,3]. miRNAs are essential for vertebrate development and are likely to be involved in differentiation and/or maintenance of tissue and cell identity. The current set of miRNAs is predicted to regulate several thousands of target mRNAs, which may include up to 30% of all protein-coding genes [4]. Therefore, there is strong interest in analysing the roles
Abbreviations: EF1a, elongation factor 1a; GAPDH, glycer-aldehyde-3-phosphate dehydrogenase; miRNA, microRNA; RT-qPCR, real-time quantitative reverse transcription polymerase chain reaction. * Corresponding author. Key Laboratory of Exploration and Utilization of Aquatic Genetic Resources, Shanghai Ocean University, Ministry of Education, Shanghai 201306, PR China. Tel.: þ86 021 61900401; fax: þ86 021 61900405. E-mail address: [email protected] (J.-L. Li). 1050-4648/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.fsi.2013.12.007
of miRNAs in development, disease and other cellular processes. miRNAs negatively regulate gene expression through two major mechanisms: translational repression and mRNA cleavage, which depend on the extent of complementarity between the miRNA and its mRNA target [5]. Since miRNAs seem to regulate gene expression by a mechanism of ‘fine-tuning’, the study of the participation of miRNAs and their targets in specific physiological or pathological experimental situations depends heavily on a reliable and accurate technique for measuring miRNA expression levels [6]. Real-time quantitative reverse transcription polymerase chain reaction (RT-qPCR) is an accurate and sensitive technique for performing gene expression analysis. The use of reference genes as internal controls for the measurement of target gene expression variation is the preferred method for normalizing RT-qPCR data because this approach captures all non-biological variation [7]. The selection of the reference gene(s) to use is not trivial and previous studies have demonstrated that a single universal reference gene is unlikely to exist and perform well for all tissue types or for all physiological, pathological and experimental situations [8,9].
X.-Y. Xu et al. / Fish & Shellfish Immunology 36 (2014) 374e382
375
Originally, normalization strategies were performed using only one reference gene; however, this idea has now evolved to include different normalization approximations, such as using the global mean normalization method or the robust multiple reference gene normalization approach where more than one reference gene is used [10]. To date, several genes, including 18S rRNA, b-actin, elongation factor 1a (EF1a) and glycer-aldehyde-3-phosphate dehydrogenase (GAPDH), have been used widely as candidate reference genes for gene expression studies in grass carp (Ctenopharyngodon idella) [11,12]. In miRNA expression studies, the most common reference genes used are ribosomal RNAs, such as 18S rRNA [13], and small nuclear RNAs like U6 snRNA [14]. However, the use of miRNAs as reference genes is not widely used at present, although it is very important that the references used have the same nature as the study subjects. The reference genes used should have the same length as the molecules of interest in order to assure the same efficiency during RNA isolation and reverse transcription [15]. In this sense, only a few studies have explored the stability of some miRNAs in human, rat and porcine tissues [6,10,16,17]. To date, no one report has deeply analysed the expression stability of miRNAs to be used as references in teleost studies. The aim of this present work was to identify suitable reference gene(s) and minimize the risk of co-regulation artifacts in grass carp. The expression of seven candidate reference miRNA genes (101a, 126-3p, 146, 192, 22a, 451 and 142a-5p) were examined in 18 developmental stages (unfertilized eggs; 0 h post-fertilization; embryos at the 16-cell stage; morula stage; gastrula stage; eye sac-appearance stage; caudal fin-appearance stage; muscular effect stage; heart beating stage; mental stages; and at 1, 2, 3, 4, 5, 6, 7 and 10 days post-hatching.) and 12 tissues (blood, brain, muscle, trunk kidney, liver, head kidney, skin, spleen, heart, gill, intestine and fin) from control fish and fish challenged with bacteria. Furthermore, the consistency of the best-scoring reference gene was tested by four statistical approaches (geNorm, BestKeeper, NormFinder and Delta CT algorithms).
total biomass. Embryos and fries were also obtained from the Wujiang National Farm of Chinese Four Family Carps, and reared in a hatching trough with constant pool water flow at 21 1 C. On day 3 post-hatching, the fry could swim steadily and were fed with freshwater rotifers captured from the pool. Twelve tissue samples were collected from three grass carp (blood, brain, muscle, trunk kidney, liver, head kidney, skin, spleen, heart, gill, intestine and fin), while embryos and early larvae were sampled at 18 different developmental stages (10 specimens from each stage). These stages were: unfertilized eggs; 0 h postfertilization; embryos at the 16-cell stage; morula stage; gastrula stage; eye sac-appearance stage; caudal fin-appearance stage; muscular effect stage; heart beating stage; mental stages; and at 1, 2, 3, 4, 5, 6, 7 and 10 days post-hatching. For the bacterial challenge, 12 fish were intrapleurally injected with formalin-killed Aeromonas hydrophila S2 (obtained from the Aquatic Pathogen Collection Centre of the Ministry of Agriculture, China) at a dose of 7.0 106 cells suspended in 100 ml PBS per fish; 12 control fish were similarly injected with 100 ml sterile PBS per fish. Three fish were sampled at 4 h, 1 day, 3 days and 7 days postinjection, respectively. Blood, brain, muscle, trunk kidney, liver, head kidney, skin, spleen, heart, gill, intestine and fin were collected from each fish. All samples were immediately snap-frozen in liquid nitrogen and stored at 80 C until use. Two groups were maintained in two aquariums and intraperitoneally injected with A. hydrophila AH10 (Aquatic Pathogen Collection Centre of Ministry of Agriculture, China) at a dose of 7.0 106 cells suspended in 100 ml PBS per fish. All the fish were observed every 4 h for any mortality and collecting samples until the termination of the experiment at 240 h post-challenge. Grass carp that died in the first 72 h post-challenge were classified as susceptible group (SG), while the animals that survived over 240 h post-challenge were considered as resistant group (RG).
2. Materials and methods
The candidate reference genes selected for evaluation included miR-148, miR-192, miR-451, miR-126-3p, miR-101a, miR-142a-5p, miR-146b and miR-22a. These miRNA candidate genes were selected due to their relative high quantities in miRNA profiles of grass carp that were generated during Illumina deep sequencing (data not shown). 18S rRNA and b-actin mRNAs were selected to be evaluated as reference genes due to their wide use in the literature. This selection process was performed after considering stability values reported in previous studies [11,12].
2.1. Sample collection Grass carp with an average weight of 50 g were cultured individually in Wujiang National Farm of Chinese Four Family Carps (Jiangsu Province, China). Animals were raised at 28 C in 400-L aerated tanks for one week before the experiment and fed twice daily (in the morning and late in the afternoon) at a ratio of 5% of
2.2. Selection of candidate reference miRNAs
Table 1 Sequence information for 14 selected candidate reference genes. Gene symbol
Primer sequence (50 e30 )
Amplification efficiency
miR-148 miR-192 miR-451 miR-let-7a miR-126-3p miR-101a miR-142a-5p miR-146b miR-22a miR-217 miR-142a-3p miR-23a B-actin (b-actin)
TCAGTGCATTACAGAACTTTGT ATGACCTATGAATTGACAGCC AAACCGTTACCATTACTGAGTT TGAGGTAGTAGGTTGTATAGTT CTCGTACCGTGAGTAATAATGC TACAGTACTGTGATAACTGAAG CATAAAGTAGACAGCACTACT TGAGAACTGAATTCCAAGGGTG AAGCTGCCAGCTGAAGAACTGT TACTGCATCAGGAACTGATTGG CGTGTAGTGTTTCCTACTTTATGGA ATCACATTGCCAGGGATTTCCA Forward: CCTTCTTGGGTATGGAATCTTG Reverse: AGAGTATTTACGCTCAGGTGGG Forward: GGACACGGAAAGGATTGACAG Reverse: CGGAGTCTCGTTCGTTATCGG
87.5 95.4 95.7 96.0 96.3 96.6 99.9 104.8 105.0 97.8 96.0 96.5 95.8
18S rRNA (18S ribosomal RNA)
98.2
376
X.-Y. Xu et al. / Fish & Shellfish Immunology 36 (2014) 374e382
Table 2 Ranking of candidate reference genes according to stability values produced by geNorm, NormFinder, BestKeeper and Delta CT, and the recommended comprehensive ranking calculated by RefFinder. A. Rank order of each miRNA gene in the different tissues of control fish. B. The geNorm stability value of each miRNA gene in the different tissues of control fish. A Ranking order (betteregoodeaverage) Method
1
2
3
4
5
6
7
8
9
Delta CT BestKeeper Normfinder geNorm Recommended comprehensive ranking
101a 22a 101a 146 j 142a-5p 101a
142a-5p 101a 142a-5p
146 18s 146 101a 146
22a 126-3p 22a 126-3p 22a
126-3p 146 126-3p 22a 126-3p
18s 142a-5p 18s 18s 18s
B-actin B-actin B-actin B-actin B-actin
451 451 451 451 451
192 192 192 192 192
142a-5p
B Gene name
146 j 142a-5p
101a
126-3p
22a
18s
B-actin
451
192
geNorm Stability value
0.87
1.222
1.503
1.639
1.761
2.108
2.358
2.835
to calculate the RT-qPCR efficiency. All standard curves were generated using 10-fold serial dilutions from a pool of cDNA from all the samples (n ¼ 30). Reactions were incubated in a 96-well plate at 95 C for 10 min, followed by 40 cycles of 95 C for 3 s and 60 C for 30 s on a CFX96 Real-Time PCR System (Bio-Rad, CA, USA). Details of all primers used are listed in Table 1. All measurements were performed in duplicate. Minus RT controls, minus poly (A) polymerase controls and no template controls were each included. Moreover, melting curve analysis was performed in each assay in order to detect non-specific amplifications.
2.3. RNA isolation and cDNA synthesis Total RNA was isolated using TRIzol reagent following the manufacturer’s recommendations (Invitrogen, Carlsbad, USA) and stored at 80 C after incubation with RNase-free gDNA Eraser (TaKaRa, Japan). The RNA concentration and purity was determined by measuring the absorbance at 260 nm and 280 nm in a Nanodrop 2000 spectrophotometer (Nanodrop Technologies, Wilmington, DE). Reversely transcribed using One Step PrimeScript miRNA cDNA Synthesis Kit (TaKaRa). This kit adds poly (A) to the 30 end of miRNA and start to reversely transcribe. The reverse transcription was led by a kind of special oligo-dT ligated with known sequence at its 50 end.
2.5. Stability expression analysis The stability of each candidate miRNA was tested using geNorm v.3.5 [9], NormFinder [18] BestKeeper [19] and the comparative delta-Ct method [20] algorithm. The geNorm algorithm calculates the gene expression stability (M) value for each candidate reference gene based on average pairwise variation between all studied genes. NormFinder is based on an analysis of variance (ANOVA) mathematical model and this estimates intra- and inter-group
2.4. RT-qPCR RT-qPCR was performed in a final volume of 20 mL and each reaction included 10 mL of SYBR Premix Ex Taq Ⅱ (2x) (Takara), 0.4 mM of Uni-miR qPCR Primer, 0.4 mM of Forward Primer and 2 mL of a 1:5 dilution of cDNA. Standard curves were generated in order
Table 3 Ranking of candidate reference genes according to stability values in different tissues of fish after bacterial challenge. The stability measurements were produced by geNorm, NormFinder, BestKeeper and Delta CT, and the recommended comprehensive ranking was calculated by RefFinder. A. Rank order of each miRNA gene in the different tissue. B. Rank order and geNorm stability value of each miRNA gene in spleen and kidney between SG and RG sample. A Ranking order (betteregoodeaverage) Tissue
1
2
3
4
5
6
7
8
9
Blood Brain Kidney Head kidney Liver Spleen Gill Heart Muscle Skin Intestine Fin
126-3p 22a 192 22a 126-3p 22a 101a 22a 22a 22a 451 22a
101a 126-3p 126-3p 142a-5p 192 101a 22a 101a 142a-5p 126-3p 142a-5p 451
451 146 22a 101a 451 126-3p 126-3p 126-3p 451 101a 22a 126-3p
22a 451 451 126-3p 22a 192 142a-5p 142a-5p 101a 146 146 18s
146 101a 146 146 101a 146 451 451 126-3p 451 192 101a
142a-5p 142a-5p 101a 18s 146 142a-5p 146 18s 18s 142a-5p 101a 142a-5p
18s 192 142a-5p 451 142a-5p 451 192 192 146 18s 126-3p 146
192 18s 18s 192 18s 18s 18s 146 B-actin 192 18s 192
B-actin B-actin B-actin B-actin B-actin B-actin B-actin B-actin 192 B-actin B-actin B-actin
B Ranking order (betteregoodeaverage) Gene name KidneygeNorm stability value Gene name SpleengeNorm stability value
126-3p 0.509 22a 0.758
192 0.535 192 0.734
146 0.572 101a 0.826
22a 0.572 126-3p 0.740
451 0.665 146 1.121
101a 0.715 142a-5p 1.166
142a-5p 1.269 451 2.300
X.-Y. Xu et al. / Fish & Shellfish Immunology 36 (2014) 374e382
377
the lowest biological variance. As no unique criterion exists to study this issue, we applied the four most frequently used programs: BestKeeper, geNorm, NormFinder and the comparative delta-Ct method. Based on the rankings from each of these programs, RefFinder calculated an overall final ranking and the results are shown in Tables 2e4. Across the different tissues examined in the miRNA studies, miR-101a was the most stably expressed gene (Table 2). The expression stability of each miRNA gene in the different RNA samples tested at various time points post-bacterial infection was evaluated by RefFinder. The most stable miRNAs were 126-3p in the blood and liver, 101a in the gill, 192 in the kidney, 451 in the intestine and 22a in the brain, head kidney, spleen, heart, muscle, skin and fin (Table 3A). Between SG and RG samples, we found same result, 192 and 126 in the kidney and 22a in the spleen was the most stable miRNAs (Table 3B). During the different developmental stages, the order of gene expression stability was 101a, 22a, 146, 142a-5p, 451, 126-3p, 18s, 192 and, finally, b-actin (Table 4). However, geNorm showed that all the miRNAs had high M values (>1.5), except for 101a and 22a. Gene with the lowest M values (
