The real time quantitative reverse transcription PCR (qRT-PCR) is becoming increasingly important to gain insight into function of genes. Given the increased sensitivity, ease and reproducibility of qRT-PCR, the requirement of suitable reference genes for normalization has become important and stringent. It is now known that the expression of internal control genes in living organism vary considerably during developmental stages and under different experimental conditions. For economically important Brassica crops, only a couple of reference genes are reported till date. In this study, expression stability of 12 candidate reference genes including ACT2, ELFA, GAPDH, TUA, UBQ9 (traditional housekeeping genes), ACP, CAC, SNF, TIPS-41, TMD, TSB and ZNF (new candidate reference genes), in a diverse set of 49 tissue samples representing different developmental stages, stress and hormone treated conditions and cultivars of Brassica juncea has been validated. For the normalization of vegetative stages the ELFA, ACT2, CAC and TIPS-41 combination would be appropriate whereas TIPS-41 along with CAC would be suitable for normalization of reproductive stages. A combination of GAPDH, TUA, TIPS-41 and CAC were identified as the most suitable reference genes for total developmental stages. In various stress and hormone treated samples, UBQ9 and TIPS-41 had the most stable expression. Across five cultivars of B. juncea, the expression of CAC and TIPS-41 did not vary significantly and were identified as the most stably expressed reference genes. This study provides comprehensive information that the new reference genes selected herein performed better than the traditional housekeeping genes. The selection of most suitable reference genes depends on the experimental conditions, and is tissue and cultivar-specific. Further, to attain accuracy in the results more than one reference genes are necessary for normalization.
Citation: Chandna R, Augustine R, Bisht NC (2012) Evaluation of Candidate Reference Genes for Gene Expression Normalization in Brassica juncea Using Real Time Quantitative RT-PCR. PLoS ONE 7(5): e36918. https://doi.org/10.1371/journal.pone.0036918
Editor: Christian Schönbach, Kyushu Institute of Technology, Japan
Received: January 6, 2012; Accepted: April 16, 2012; Published: May 11, 2012
Copyright: © 2012 Chandna et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The study was financially supported by the core grant from National Institute of Plant Genome Research (NIPGR), New Delhi, India. RC was supported by the Research Associate fellowship from NIPGR. RA was funded by a fellowship from Council of Scientific and Industrial Research (CSIR), India. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Gene expression analysis is extremely important in many fields of biological research. Understanding the expression pattern of genes provides a useful mean of studying the complex regulatory networks occurring in living organism. Among the widely used methods to measure the levels of gene expression, real time quantitative reverse transcription PCR (qRT-PCR) represents a suitable technology . Being an efficient, sensitive, and reliable method, qRT-PCR provides a rapid mean towards simultaneous measurement of gene expression across different samples . Since this platform is relatively simple coupled with a high level of sensitivity, qRT-PCR is rapidly being adopted as a standard method for performing in-depth expression analysis of number of target genes. For accurate and reliable analysis of target gene expression, normalization of qRT-PCR data with suitable internal reference gene(s) is required . Normalization is essential to correct the non-specific variations arising because of the difference in amount of template used and its quality that can affect the efficiency of the qRT-PCR reactions . Normalization also allows the direct comparison of normalised transcript expression levels between samples . An ideal reference gene should express at constant level in all tissues and at all developmental stages, regardless of the experimental conditions or treatments , .
Commonly used reference genes are mostly cellular maintenance genes (also known as housekeeping genes), which are involved in basic and ubiquitous cellular processes such as components of the cytoskeleton, glycolytic pathway, protein folding, protein degradation, synthesis of ribosome subunits. Most frequently used housekeeping genes including β-actin (ACT), α-tubulin (TUA), ubiquitin (UBQ), glyceraldehde-3-phosphate dehydrogense (GAPDH), 18S or 26S ribosomal RNA and elongation factors (EF) have been validated as suitable internal control genes in many plants , –. These genes were assumed to be expressed constitutively and also have constant expression levels between different samples. However, there are evidences that transcripts levels of housekeeping genes vary considerably across the developmental stages and under variable conditions . This variability in expression across experimental samples may be because these housekeeping genes not only participate in basic cell metabolism but also in other cellular processes . Therefore, selecting multiple stably expressed reference genes, other than the commonly used housekeeping genes need to be considered for accurate normalization of gene expression studies .
Recognising the importance of validation of reference gene(s) for normalization of qRT-PCR data initiated the development of number of software packages such as geNorm ,  and NormFinder  Usage of these statistical algorithms have greatly simplified the selection of appropriate reference genes by calculating the expression stability and determining the optimal number of candidate reference genes required for normalization under specific conditions in various organisms, including plants , . A number of attempts for reference gene validation have been reported in plants such as rice , chickpea , potato , soybean , , tomato , chrysanthemum , grape , cabbage , wheat , Brassica napus  and poplar .
The Brassica species have diverse characteristics and are of great agronomic importance as vegetables, condiments, fodder, and oil crops. Brassica crops are globally the third most important sources of vegetable oil after soybean and groundnut . Brassica juncea (brown or Indian mustard) is an important oilseed crop cultivated mainly in Indian sub-continent besides some parts of east Europe, Africa, Canada and China. Only a limited number of gene expression studies have been carried out in B. juncea, wherein ACT and 18S are the commonly used reference genes –.
Comparison of several candidate reference genes in Brassica crops, particularly in B. juncea, is not yet reported, thereby limiting our knowledge about the choice of best reference gene which could be used for normalization of gene expression across developmental stages and variable growth conditions. In present study, we have compared the performance of 12 candidate reference genes (consisting of five commonly used housekeeping genes of plants, and seven new candidate reference genes selected from B. rapa and B. napus microarray platforms) in 49 diverse samples of B. juncea, broadly categorized into five distinct experimental sets. Our results reveal that new reference genes are more stably expressed than the traditionally used housekeeping genes across all the five experimental sets. Further, combination of most stable reference genes provides a more accurate and reliable mean of normalization during qRT-PCR analysis.
Selection of Candidate Reference Genes and Primer Design
A total of 12 candidate reference genes, including five traditional housekeeping genes namely, actin2 (ACT2), elongation factor 1B (ELFA), glyceraldehyde-3-phosphate- dehydrogenase (GADPH), α-tubulin (TUA) and ubiquitin 9 (UBQ9) and seven new reference genes, acyl carrier proteins (ACP), clathrin adaptor complex (CAC), sucrose non fermenting-1 protein kinase (SNF), tonoplastic intrinsic proteins-41 (TIPS-41), trans membrane proteins (TMD), tryptophan synthase-β (TSB) and zinc finger protein (ZNF) were used in this study. The new candidate reference genes were selected on the basis of their stable expression profiles across developmental stages and during abiotic stress conditions, as determined using microarray data of B. rapa (www.brassica.info/resource/trancriptomics.php)  and B. napus (www.rapeseed.plantsignal.cn) ,  and analysed using A. thaliana gene expression tool (www.jsp.weigelworld.org)  (File S1, S2 and S3).
Since, sequence information of B. juncea is very limited, we used publicly available gene sequences from Arabidopsis and related Brassica species to design the gene specific primers. The Brassica specific expressed sequence tags (EST) and genome survey sequences (GSS) of these candidate reference genes were largely obtained from the Brassica genome gateway (http://brassica.bbsrc.ac.uk)  and the recently available B. rapa genome portal (http://brassicadb.org/brad) , by providing Arabidopsis complementary DNA sequence (CDS) as a query. To ensure gene amplification in B. juncea, the Arabidopsis and Brassica sequences of each candidate gene were aligned together and the primers were designed from the consensus regions of the aligned sequences, preferably spanning an intron (File S4).
Verification of Primer Specificity and PCR Efficiency Analysis
In order to determine specificity of primers designed in the current study, agarose gel electrophoresis and melting curve analyses were performed following the qRT-PCR experiment on seedling stage of B. juncea L. cv. Varuna. All the primer pairs amplified single PCR product of expected size (File S5) and the specificity of amplicon was confirmed by the presence of single peak during melt curve and sequencing analysis (File S6 and S7). A standard curve was generated using 10-fold serial dilutions of cDNA to calculate the gene specific PCR efficiency. The slopes of the standard curves were used to calculate the correlation coefficient (R2) and PCR efficiency (File S8). The linear R2 for all the primers ranged between 0.994–0.999 over 1000 fold of cDNA dilution. Further, PCR efficiencies of primers ranged from 94%–106% (Table 1).
Expression Profiling of B. juncea Reference Genes
A real-time qRT-PCR assay, based on SYBR Green detection, was designed for transcript profiling of the 12 candidate reference genes (ACP, ACT2, CAC, ELFA, GAPDH, SNF, TIPS-41, TMD, TSB, TUA, UBQ9 and ZNF) in 49 diverse samples of B. juncea (Table 1). In order to minimize the variability associated with qRT-PCR analysis, all RNA samples were adjusted to same concentration and quality pass prior to their conversion into cDNA (File S9). The expression level of the candidate genes obtained during qRT-PCR experiments are presented as threshold cycle (Ct) values. The 12 reference genes used in the current study showed relatively wide ranges of Ct values across 49 samples of B. juncea. The mean Ct values of reference genes ranged from 22.02–29.36 (Figure 1). GAPDH was the most abundant reference gene of the set (mean Ct = 22.02), whereas TMD was the least abundant reference gene (mean Ct = 29.36). Interestingly, three new candidate reference genes CAC, TIPS-41 and SNF showed least expression variation (Co-variance, CV of 4.34, 4.96 and 4.98, respectively); while the commonly used housekeeping genes GAPDH and ELFA (CV of 17.77 and 14.57, respectively) had the most variable expression profile across all the 49 experimental samples (Figure 1).
The boxes represent mean Ct values and bars correspond to the standard deviation. The Ct values were calculated on 1∶50 diluted cDNA samples.
In order to perform in-depth expression analysis of candidate reference genes, the 49 diverse samples were categorized under five experimental sets comprising of defined development or condition-specific samples of B. juncea (Table 2). The first experimental set comprised of six vegetative stages (cotyledon, seedling, young leaf, mature leaf, stem and roots), whereas the second set included six reproductive stages (flower bud, flower, pod 5 days post anthesis (dpa), pod 10 dpa, pod 15 dpa and pod 30 dpa). In the third experimental set, all the 12 developmental stages of B. juncea (both vegetative and reproductive stages as mentioned above) were analysed together. The fourth set consisted of 12 samples treated with different abiotic stress conditions and hormones. The fifth experimental set included the vegetative stages of Indian and east-European B. juncea cultivars (Pusa Bold, Kranti, Donskaja, Early Heera-2, and Zem 84500).
It was observed that the tissue type and experimental conditions does affect the expression of reference genes. In vegetative stages, TMD, ELFA and ACT2 showed the least expression variation whereas a large variation in the expression levels of UBQ9 and ZNF was observed (Table 2). In reproductive stages TSB, ACT2 and ELFA showed large variation in their expression profiles, whereas CAC and GAPDH were the least variable transcripts. Interestingly, across total developmental stages, the two new reference genes TIPS-41 and CAC emerged as most stable transcripts showing least variation and UBQ9 and ZNF were the most variable transcripts (Table 2). In contrast, the expression of UBQ9 was least variable in stress and hormone treated samples, whereas expression of GAPDH varied widely in various treated samples (Table 2). A significant variation in the expression of ELFA and GAPDH was also observed across the cultivars, while CAC and TIPS-41 showed least expression variation (Table 2). In general, expression of all the 12 reference genes selected herein showed a large variation across the five cultivars of B. juncea. This probably reflects the presence of allelic variation in these genes across the B. juncea cultivars. Thus, transcript levels of none of the candidate reference genes remain constant throughout the developmental stages, stress and hormone treatments and across the B. juncea cultivars.
Gene Expression Stability Analysis and Ranking of B. juncea Reference Genes
The expression stability of 12 candidate reference genes, across five experimental sets, was measured and ranked using two different programs namely geNorm  and NormFinder . Based on geNorm analysis, ELFA, ACT2, CAC, and TIPS-41 (in order) were identified as the most stable genes across the vegetative stages whereas TIPS-41, CAC, GAPDH and TUA remained the most stable genes across the reproductive stages (Figure 2). When all the 12 samples were analysed together, GAPDH, TUA, TIPS-41 and CAC were the most stable genes, while UBQ9 and ZNF were least stable transcripts, in order. It is interesting to note that most stable housekeeping genes of vegetative stage (ELFA and ACT2), were not among the most stable reference genes of the total developmental set in B. juncea. Further, in response to various stress and hormone treatments, UBQ9, TIPS-41, ZNF and CAC were the most stable genes whereas across the B. juncea cultivars, TIPS-41, CAC, ZNF and TUA were identified as the most stable genes, in order. The commonly used housekeeping gene GAPDH was, however, found to be among the least stable transcript in treated samples as well as across the cultivars of B. juncea (Figure 2).
Mean expression stability (M) following stepwise exclusion of the least stable gene across all the samples within an experimental set. The least stable genes are on the left, and the most stable on the right.
The NormFinder also ranked ELFA, TIPS-41, CAC and ACT2 as the most stable reference genes in vegetative stages (in order), whereas in reproductive stages TIPS-41, CAC, ZNF and TMD were identified as the most stable reference genes (Table 3). Across total developmental stages GAPDH, TIPS-41, ELFA and TUA were the most stable genes while ZNF and UBQ9 were the least stable genes (Table 3). Both, CAC and TIPS-41 also emerged as most stable reference genes in stress and hormone treated samples and also across the B. juncea cultivars. Thus using both programs, similar genes were identified as stable reference genes across all the five experimental sets although the gene expression stability rankings for some reference genes were slightly altered.
Optimal Number of Reference Gene for Normalization Across the Experimental Sets
The geNorm software was further used to calculate the optimal number of reference genes necessary for normalization across different sets of experiment. The pairwise variation (Vn/Vn+1) between sequential normalization factors, NFn and NFn+1 was used to determine the number of genes required for reliable normalization . As shown in Figure 3, differences in the expression stability values of the candidate reference genes were less marked in reproductive stages, than in other series (Figure 3). The V2/3 value for reproductive stage was 0.108 (geNorm V <0.15 when comparing a normalization factor based on the two or three most stable targets), so CAC together with TIPS-41 would be sufficient for normalization purpose in reproductive stages (Figure 3). In vegetative stages, V4/5 value was 0.132 thereby suggesting that the optimal number of reference targets would be four, namely ELFA, ACT2, CAC and TIPS-41. However on analysing all the developmental stages together, four reference genes namely GAPDH, TUA, TIPS-41 and CAC should be considered (V4/5 value = 0.137). In treated samples, the pair of TIPS-41 and UBQ9 produced a V2/3 value of 0.12, therefore these two candidate reference genes can be used for normalization for a wide range of stressed and hormone treated tissue samples in B. juncea. Across cultivars, TIPS-41 and CAC could be the choice of optimal reference genes (V2/3 value of 0.124). Thus across all the five experimental sets of B. juncea, the two new candidate reference genes namely TIPS-41 and CAC were identified as the most suitable reference genes for normalization in gene expression studies.
Reference Gene Validation
To validate the selection of candidate reference genes for normalization, we analyzed the normalized fold expression of BjDREB2 (a dehydration responsive element binding proteins-2 homolog of B. juncea). DREB2 is a transcription factor that imparts stress endurance to plants and plays a crucial role in providing tolerance to heat, dehydration, wounding and salt stresses. UBQ9, TIPS-41 and ZNF reference genes identified as the most stable genes using geNorm analysis among all stress treated samples, were further tested for normalization. The transcript abundance of BjDREB2 increased in dehydration stressed sample when normalized using all the three genes independently, although at different level (Figure 4). The expression pattern of BjDREB2 transcript was higher in heat stress in case of UBQ9, whereas the increased expression pattern was observed in dehydration samples using either ZNF or TIPS-41 as the reference gene. However, when the expression of BjDREB2 was normalized using a combination of UBQ9 and TIPS-41, identified by geNorm as most stable reference genes, the fold expression of BjDREB2 was highest in dehydration stressed sample. This data clearly suggests that the use of more than one reference genes for normalization provides more accurate representation of target gene expression tested across the variable experimental conditions. As expected, the BjDREB2 expression remains unaltered in hormone treated samples (data not shown).
Quantitative real time PCR has become a powerful technique for gene expression studies, because of its high throughput, sensitivity and accuracy . The choice of stably expressed reference genes for normalization is the paramount to accurate interpretation of the results. The normalization takes care of the variation introduced by the quantity and quality of input RNA, its cDNA conversion and the various steps involved in the qRT-PCR assay. Normalization studies with multiple reference genes validated for their expression stability is required for reliable gene expression results, as no single reference gene in plants has been shown to have a stable expression during variable experimental conditions including different samples/treatments. The usage of statistical algorithms, such as geNorm and NormFinder, has greatly simplified the selection and validation of reference genes by calculating the expression stability of reference genes as well as determining the number of reference genes required for accurate normalization across the experimental conditions tested , .
This study describes a comprehensive analysis on the validation of 12 candidate reference genes (including five commonly used housekeeping genes of plants and seven new candidate reference genes) in 49 diverse samples of B. juncea, divided broadly into five experimental sets. Our analysis based on geNorm and NormFinder algorithms indicated that the choice of reference genes for normalization should be experiment and stage-specific. For example, across total developmental stages of B. juncea (including both vegetative and reproductive stages) four genes namely GAPDH, TUA, TIPS-41 and CAC are ideal for normalization factor. However, when the total developmental stages were subdivided into two distinct experimental sets viz., the vegetative and reproductive stages, different sets of candidate reference genes appeared to be the best for normalization in each stage (Figure 2; Table 3). For vegetative stages, the four reference genes viz., ELFA, ACT2, CAC and TIPS-41 are ideal for calculating normalization factor, however for reproductive stages TIPS-41 and CAC are enough for calculating the normalization factor. Since seed development is an economically important and distinct phase in Brassica crops, identification of suitable reference genes, particularly in reproductive stages will greatly assist in performing in-depth expression analysis of seed-specific genes in the said crop plants.
Some of the novel candidate reference genes selected in the current study performed better than the traditional housekeeping genes under each experimental set. TIPS-41 (Tonoplast intrinsic proteins) and CAC (Clathrin adaptor complex), were identified amongst the list of most stable reference genes in all the five experimental sets of B. juncea tested in this study. TIPS-41 (or TIP41) is one of the top ranked reference genes identified in the both vegetative and reproductive stages of B. juncea and B. napus, probably reflecting its similar and stable transcription regulation across the Brassica crops . Previous studies in selection of reference genes during tomato development, Fagopyrum esculentum developmental stages and soybean under varied light regimes , ,  also identified TIPS-41 as the most stable reference gene for the vegetative sample. Similarly, CAC has also proved to be the best candidate for normalization in banana fruit, tomato, coffee, buckwheat and Cucurbita pepo studies , –.
In current study, we tested the expression stability of commonly used housekeeping genes like GAPDH, UBQ, ACT, ELFA and TUA that have been previously described as “candidate controls” in various plant studies. Some of these housekeeping genes qualify among the best reference genes under specific experimental sets of B. juncea, however none of them was found to be the suitable reference gene across all the five experimental sets of B. juncea. For example, UBQ9 was found to be the most appropriate gene in stress and hormone treated samples while expression stability of UBQ9 ranked late in different sets of samples including total developmental stages (Figure 2; Table 3). Our findings are in accordance with the previous studies, wherein UBQ9 and its putative homologues were the least stably expressed genes during total developmental stages in rice, soybean, grapeberry, zucchini and chicory , –, ,  Expression of homologues of UBQ was also found to be stable when evaluated in chrysanthemum and banana under various stresses , .
Housekeeping genes like GAPDH and ELFA were also not found to be expressed stably across the diverse experimental sets of B. juncea, tested in the current study. The expression of GAPDH and ELFA was only found to be stable in total development and vegetative stages, respectively (Figure 3; Table 3). Earlier studies have shown that GAPDH has constant expression in total developmental stages of chickpea and sugarcane, while it was observed as the least stable transcript in wheat developmental series , , . Similarly, ELFA was also reported as one of the best reference control gene across vegetative stages of rice, perennial ryegrass, chickpea, Linum usitatissimum, Brachypodium distachyon, Arabidopsis and chicory , , , , –. However, there are evidences that the expression profile of ELFA was not as consistent as that of other reference genes in soybean, Salvia miltiorrhiza and tomato , , . The poor performance of ACT2 across total development, reproductive stages, treatments and cultivars of B. juncea was surprising since this gene has been used as a reference control in earlier gene expression studies –. However, ACT2 can be used in combination of other selected genes for normalization in vegetative stages. Recently, several studies have also shown that the use of ACT for normalization is not reliable in rice , potato , Arabidopsis  and peach .
The varied expression profiles of commonly used housekeeping genes may be because they are reported to be involved in many other cellular processes besides their basic cellular metabolic functions. For example, GAPDH not only acts as a component of the glycolytic pathway, but it is also involved in other processes such as cell proliferation . Similarly, ubiquitin which primarily participates in proteolytic degradation, also has non-proteolytic functions . The varied expression of actin may be due to its participation in cytoplasmic streaming, cell division and the distribution of the plasma membrane proteins other than being a major component of eukaryotic cytoplasmic microfilaments .
The newly selected reference genes of B. juncea, are superior to traditional ones in terms of their expression stability. The microarray expression data of B. rapa and B. napus also showed better expression stability of new reference genes identified, compared to the traditional ones (File S1 and S2). Further, the findings of this study are also in good accordance with the Arabidopsis microarray expression data of these 12 candidate reference genes available at AtGenExpress Visualization Tool portal , which compiles the expression profiles of five datasets of Arabidopsis (developmental, hormones, abiotic stress, light and pathogen; FileS3). Finally, using the gene expression profiling of BjDREB2 in abiotic stressed samples of B. juncea, we found that normalization involving the combination of more than one stable reference genes resulted in improved accuracy (Figure 4).
In the present study, we evaluated the expression stability of 12 candidate reference genes across large number of B. juncea samples in an effort to identify a set of stable reference gene(s) for normalization during gene expression studies. Analysis of expression stability using geNorm and NormFinder revealed that the expression of TIPS-41 and CAC are most stable across variable experimental tissues. In addition, data analysis using geNorm suggested that three housekeeping genes (ELFA, UBQ9 and GAPDH) can be used in combination with CAC and/or TIPS-41 to calculate the normalization factor based on multiple reference genes. Although no candidate reference gene was constantly superior to the others, our data suggest that the novel genes performed better that commonly used housekeeping genes of B. juncea. We conclude that the results outline in the present study will facilitate sensitive and accurate quantification of gene expression in B. juncea which could also be extrapolated to related Brassica crops. Further, in the absence of enriched genome and transcriptome information from B. juncea and its diploid progenitor parents, the current study will greatly assist the Brassica research community to select a set of novel reference genes which could potentially be used for large arrays of experimental conditions and more-importantly in a cross-species manner.
Mustard (Brassica juncea L. cv. Varuna) was used for the experiments. A total of 12 tissues including six vegetative stages (cotyledons, seedlings, young leaf (20 days post sowing), mature leaf (40 dps), root and stem) and six reproductive stages (bud, flower, pod 5 dpa (days post anthesis), pod 10 dpa, pod 15 dpa and pod 30 dpa) were collected from the plant growing in field condition.
Five B. juncea cultivars were also included in the study: Early Heera2 (EH2), Pusa Bold (PB), Kranti (KR), Donskaja (DK) and Zem 84500 (ZM). Five vegetative stages (cotyledons, seedlings, leaf, stem and roots) were harvested from these cultivars.
Stress treatments: For stress treatments, 7-day old seedlings were used. Seeds were grown in half strength Murashige-Skoog (MS) media till five days. Elicitors were added after adapting 6-day old seedlings in sucrose free liquid medium for 24 h in dark. Thereafter, seedlings were transferred to beakers containing MS along with sodium chloride (NaCl, 300 mM), abscisic acid (ABA, 100 µM), methyl jasmonate (MeJa, 200 µM), salicylic acid (SA, 200 µM), glucose (3%), indole-3-acetic acid (IAA, 100 µM) and 1-aminocyclopropane-1-carboxylate deaminase (ACC,100 µM), and incubated for 6 h.
For drought treatment seedlings were air dried for 6 h duration; for cold and heat shock treatments the seedlings were kept at 4±1°C and 42±1°C respectively, for 6 h. For wound treatment seedlings were wounded with blunt forceps and collected after 10 min. The mock treated seedling for same interval served as control.
Totally, the experimental samples comprised of 12 developmental stages, 12 exposed to various stress treatments and 25 samples of vegetative stages involving five different cultivars, thereby consisting a total of 49 different tissues.
RNA Isolation, Quality Control and cDNA Synthesis
RNA was extracted from all the plant tissues using the Spectrum Plant Total RNA Kit (Sigma Life Sciences, USA) according to manufacturer’s instructions. The quantity and quality of RNA sample was checked using Nano spectrophotometer (ND-1000 Thermo scientific); and all RNA samples were adjusted to the same concentration. RNA quality was further assessed using the Agilent-2100 Bioanalyzer and RNA 6000 Nano chips (Agilent Technologies, Singapore). RNA samples with 260/280 ratio from 1.9 to 2.1, 260/230 ratio from 2.0–2.5 and RIN (RNA integrity number) more than 7, were used for further analysis (File S9). The integrity of RNA samples were also checked by agarose gel electrophoresis. For each method the measurement was done in duplicates.
First strand cDNA was synthesized by reverse transcribing 2 µg of total RNA with high-capacity cDNA Reverse Transcription kit (Applied Biosystems, USA) in a 20 µl reaction using mixture of random primers and oligo-dT’s in 1∶1 ratio according to manufacturer’s instructions. cDNA was diluted 50 times for the use of real-time qRT-PCR reaction. All cDNA were stored at −20°C until PCR.
Selection of Reference Genes and Primer Designing
The five traditional housekeeping genes often used as references control in plants were selected: ACT (Actin2, At3G46520), UBC9 (Ubiquitin9, At4G27960), ELFA (Elongation factor 1B, At1G09640), TUA (α-Tubulin, At5G19770), GAPDH (Glyceraldehyde-3-phospho dehydrogense, At3G04120).
In addition, seven new reference control genes were selected using the available B. rapa www.brassica.info/resource/trancriptomics.php  and B. napus www.rapeseed.plantsignal.cn ,  microarray experiments, covering various abiotic stresses and developmental stages. The ratio of expression levels between control and treatment experiments for each gene that was within the limit of two-fold were selected as putative candidate reference genes. Based on the stable expression profiles during abiotic stress and development stages in B. rapa microarray transcriptome data, sequences of four potential new reference genes: ACP (Acyl carrier proteins, Arabidopsis ortholog At1G54630), SNF (Sucrose non fermenting-1 protein kinase, At3G50500), TMD (Leucine rich trans membrane domain proteins, At5G22600), TIPS-41 (Tonoplastic intrinsic proteins, At4G34270) were selected (File S1). The three remaining genes, CAC (Clathrin adaptor complex, At5G46630), TSB (Tryptophan synthase-β, At4G27070) and ZNF (Zinc finger protein, At1G01930) were selected on the basis of non-significant change in their expression profiles during seed development in B. napus transcriptome analysis (File S2).
The full length complementary DNA (CDS) of the Arabidopsis genes were used to query homologous Brassica sequences. Expressed sequence tags (ESTs) and Genome survey sequences (GSS) were obtained from the publically available platform at NCBI, Brassica genome gateway http://brassica.bbrc.ac.uk  and recently available B. rapa genome sequence portal http://www.brassica-rapa.org . The primer for DREB2 (dehydration responsive binding element) were designed using CDS of Arabidopsis DREB2 gene and its B. rapa gene orthologous (Bra005852 and Bra009112; available on Brassica database http://brassicadb.org . MegAlign module of DNASTAR was used to align the sequenced Brassica EST’s/GSS’s with the known Arabidopsis CDS. The qRT-PCR primer was designed from the consensus sequences preferably spanning the intron(s) (File S4). The details of primer sequences are given in Table 1.
For all genes, primer pairs were designed using the online available Sigma DNA Calculator with the following parameters: optimal length 20–25 nucleotides, melting temperature 60–65°C, GC content <50%, product size range 80–200 base pairs, minimum or no self complementarities at 3′ end, absence for the hairpin structures and self-dimers. In order to confirm the sequences of the amplicons, PCR was performed on cDNA for all designed primer pairs. The products were analyzed on 2% agarose gel and sequenced (File S7). A series of 10 fold of three dilutions of cDNA (10–1,000 fold dilution), were made to determine the gene specific PCR amplification efficiency for each primer pair in qRT-PCR experiments. Based on the Ct values for all dilution points in a series, a standard curve was generated using linear regression and the slope. The qbase http://medgen.ugent.be/~jvdesomp/genorm  calculated the gene specific PCR amplification efficiency of the primer using the following equation: Efficiency % = 10(−1/slope)×100%.
Mean of expression levels for all the 12 genes studied in 49 tissue samples were calculated. Standard deviation was calculated using Microsoft Excel and Co-variance was calculated as Standard Deviation/Mean ×100.
Real-time qRT-PCR Assays
Real-Time PCR was performed in an optical 96-well plate with an 7900 HT real time PCR machine (Applied Biosystems) and universal cycling conditions (95°C for 10 min, 40 cycles of 15 s at 95°C and 60°C for 60 s) in final volume of 20 µl. Reactions contained SYBR Green Master Mix (Kapa Biosystems), 10 pM of a gene specific forward and reverse primers and 2 µl of the diluted cDNA. A no template control (NTC) was also included in each run for each gene; in this study 2 µl RNase free water was used. Each experiment was conducted in three technical replicates with at least two biological replicates for each tissue. To check for the specificity of PCR amplification dissociation curve was generated. The Ct values were automatically determined for each reaction using SDS version 2.3 and RQ manager version 1.2 (Applied Biosciences) software with default parameters.
Following PCR data collection, two publically available software tools, called geNorm v3.2  and NormFinder v0.953  were used to rank the expression stability of reference genes across all the five experimental sets. The procedures outlined in the user’s manuals, http://medgen.ugent.be/~jvdesomp/genorm  and http://www.mdl.dk/publicationsnormfinder.htm , were followed to calculate instability values of gene expression.
Briefly, the geNorm program is based on pairwise comparisons and stepwise exclusion of candidate genes according to their expression stability measures (M) values. In general, lower the M value, higher the gene expression stability. geNORM recommends M<1.5 to identify sets of reference genes with stable expression. Further, the pairwise variation (Vn/Vn+1) between sequential normalization factors, NFn and NFn+1 was used to determine the number of genes required for reliable normalization. A threshold value of 0.15 was reported by Vandesompele et al.  for normalization of expression stability. It has been suggested that below this threshold values there is no need of an additional internal control gene.
The NormFinder reference tool was also applied to rank the candidate reference genes expression stability for all the samples with no subgroup determination. NormFinder used ANOVA based model to estimate intra- and inter-group variation, and combines these estimates to provide a direct measure of the variations in expression for each gene . It ranks the genes according to their stability under given set of experimental conditions. Genes with lower average expression stability values are more stable.
Gene expression profile of the genes considered in present study, as observed in B. rapa microarray database (http://www.brassica-rapa.org).
Arabidopsis microarray expression data of development stages showing the expression profile of 12 candidate reference genes considered in this study (available at jsp.weigelworld.org).
An example of DNA sequence alignment of Arabidopsis CDS and its homologous Brassica EST and GSS. The nucleotide bases highlighted in grey are the conserved sequences for designing primers for TIPS-41.
The confirmation of expected amplicon size of the primer pairs.
Melt curves of the candidate reference genes under different experimental conditions.
Sequencing data of the PCR amplicons.
Amplification efficiencies of the primers designed in the current study.
The central instrumentation facility at National Institute of Plant Genome Research (NIPGR) is acknowledged. We thank Dr. Manoj Majee for his critical comments on the manuscript.
Conceived and designed the experiments: RC NCB. Performed the experiments: RC RA NCB. Analyzed the data: RC NCB. Contributed reagents/materials/analysis tools: NCB. Wrote the paper: RC NCB.
- 1. Bustin SA, Benes V, Nolan T, Pfaffl MW (2005) Quantitative real-time RT-PCR – a perspective. J Mol Endocrinol 34: 597–601.
- 2. Bustin SA, Benes V, Garson JA, Hellemans J, Huggett J, et al. (2009) The MIQE guidelines: minimum information for publication of quantitative real-time PCR experiments. Clin Chem 55: 611–622.
- 3. Czechowski T, Stitt M, Altmann T, Udvardi MK, Scheible WR (2005) Genome wide identification and testing of superior reference genes for transcript normalization in Arabidopsis. Gen Anal 139: 5–17.
- 4. Andersen CL, Jensen JL, Orntoft TF (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.
- 5. Maroufi A, Bockstaele EV, Loose MD (2010) Validation of reference genes for gene expression analysis in chicory (Cichorium intybus) using quantitative real-time PCR. BMC Mol Biol 11: 15–27.
- 6. Vandesompele J, De Preter K, Pattyn F, Poppe B, Van Roy N, et al. (2002) Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol 3: 1–11.
- 7. Radonic A, Thulke S, Mackay IM, Landt O, Siegert W, et al. (2004) Guideline to reference gene selection for quantitative realtime PCR. Biochem Biophys Res Commun 313: 856–862.
- 8. Nicot N, Hausman JF, 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.
- 9. Hu R, Fan C, Li H, Zhang Q, Fu Y-F (2009) Evaluation of putative reference genes for gene expression normalization in soybean by quantitative real-time RT-PCR. BMC Mol Biol 10: 93–104.
- 10. Garg R, Sahoo A, Tyagi AK, Jain M (2010) Validation of internal control genes for quantitative gene expression studies in chickpea. Biochem Biophys Res Commun 396: 283–288.
- 11. Gu C, Chen S, Liu Z, Shan H, Luo H, et al. (2011) Reference gene selection for quantitative real-time PCR in Chrysanthemum subjected to biotic and abiotic stress. Mol Biotechnol 49: 192–197.
- 12. Guenin S, Mauriat M, Pelloux J, Van Wuytswinkel O, Bellini C, et al. (2009) Normalization of qRT-PCR data: the necessity of adopting a systematic, experimental conditions-specific, validation of references. J Exp Bot 60: 487–493.
- 13. Hellemans J, Mortier G, De Paepe A, Speleman F, Vandesompele J (2007) qBase relative quantification framework and software for management and automated analysis of real-time quantitative PCR data. Genome Biol 8: R23.
- 14. Brunner AM, Yakovlev IA, Strauss SH (2004) Validating internal controls for quantitative plant gene expression studies. BMC Plant Biol 13: 4–14.
- 15. Dheda K, Huggett JF, Chang JS, Kima LU, Bustin SA, et al. (2005) The implications of using an inappropriate reference gene for real-time reverse transcription PCR data normalization. Anal Biochem 37: 141–143.
- 16. Jain M, Nijhawan A, Tyagi AK, Khurana JP (2006) Validation of housekeeping genes as internal control for studying gene expression in rice by quantitative real-time PCR. Biochem Biophy Res Commun 345: 646–651.
- 17. Jian B, Liu B, Bi Y, Hou W, Wu C, et al. (2008) Validation of internal control for gene expression study in soybean by quantitative real-time PCR. BMC Mol Biol 9: 59–72.
- 18. Expósito-Rodrı ´guez M, Borges A, Borges-Pérez A, Pérez J (2008) Selection of internal control genes for quantitative real-time RT-PCR studies during tomato development process. BMC Plant Biol 8: 131–142.
- 19. Reid KE, Olsson N, Schlosser J, Peng F, Lund ST (2006) An optimized grapevine RNA isolation procedure and statistical determination of reference genes for real-time RT-PCR during berry development. BMC Plant Biol 6: 27–37.
- 20. Qi J, Yu S, Zhang F, Shen X, Zhao X, et al. (2010) Reference Gene Selection for Real-Time Quantitative Polymerase Chain Reaction of mRNA Transcript Levels in Chinese Cabbage (Brassica rapa L. ssp. pekinensis). Plant Mol Biol Rep 28: 597–604.
- 21. Long X, Wang JR, Ouellet T, Rocheleau H, Wei Y, et al. (2010) Genome-wide identification and evaluation of novel internal control genes for Q-PCR based transcript normalization in wheat. Plant Mol Bio 74: 307–311.
- 22. Chen X, Truksa M, Shah S, Weselake RJ (2010) A survey of quantitative real-time polymerase chain reaction internal reference genes for expression studies in Brassica napus. Anal Biochem 405: 138–140.
- 23. Gutierrez L, Mauriat M, Guénin S, Pelloux J, Lefebvre JF, et al. (2008) The lack of a systematic validation of reference genes: a serious pitfall undervalued in reverse transcription polymerase chain reaction (RT-PCR) analysis in plants. Plant Biotechnol J 6: 609–618.
- 24. A.O F (2009) Food and Agriculture Organization of the United Nations.
- 25. Muthukumar B, Yakubov B, Salt DE (2007) Transcriptional activation and localization of expression of Brassica juncea putative metal transport protein BjMTP1. BMC Plant Bio 7: 32–44.
- 26. Jami SK, Clark GB, Turlapati SA, Handley C, Roux SJ, et al. (2008) Ectopic expression of an annexin from Brassica juncea confers tolerance to abiotic and biotic stress treatments in transgenic tobacco. Plant Physiol Biochem 46: 1019–1030.
- 27. Srivastava S, Srivastava AK, Suprasanna P, D’Souza SF (2009) Comparative biochemical and transcriptional profiling of two contrasting varieties of Brassica juncea L. in response to arsenic exposure reveals mechanisms of stress perception and tolerance. J Exp Bot 60: 3419–3431.
- 28. Xu J, Chai T, Zhang Y, Lang M, Han L (2009) The cation-efflux transporter BjCET2 mediates zinc and cadmium accumulation in Brassica juncea L. Leaves. Plant Cell Rep 28: 1235–1242.
- 29. Kaur P, Jost R, Sivasithamparam K, Barbetti MJ (2011) Proteome analysis of the Albugo candida–Brassica juncea pathosystem reveals that the timing of the expression of defence-related genes is a crucial determinant of pathogenesis. J Exp Bot 62: 1285–1298.
- 30. Brassica rapa transcriptome database. 20: Available: http://www.brassica.info/resource/trancriptomics.php. Accessed 2011 Dec.
- 31. Niu Y, Wu GZ, Lin WH, Shi QM, Xue LJ, et al. (2009) Global analysis of gene expression profiles in Brassica napus developing seeds reveals a conserved lipid metabolism regulation with Arabidopsis thaliana. Mol Plant 2: 1107–1122.
- 32. Brassica napus transcritome database. 20: Available: http://www.rapeseed.plantsignal.cn. Accessed 2011 Dec.
- 33. AtGenExpress visualization tool. 30: Available: http://jsp.weigelworld.org Accessed 2011 Nov.
- 34. Brassica genome gateway. 28: Available: http://brassica.bbsrc.ac.uk. Accessed 2011 Nov.
- 35. B. rapa genome portal. 5: Available: http://brassicadb.org/brad. Accessed 2011 Dec.
- 36. Demidenko NV, Logacheva MD, Penin AA (2011) Selection and validation of reference genes for quantitative real-time PCR in Buck wheat (Fagopyrum esculentum) based on transcriptome sequence data. PLoS One 6: e19434.
- 37. Cruz F, Kalaoun S, Nobile P, Colombo C, Almeida J, et al. (2009) Evaluation of coffee reference genes for relative expression studies by quantitative real time RT-PCR. Mol Breed 23: 607–616.
- 38. Chen L, Zhong H, Kuang J, Li J, Lu W, et al. (2011) Validation of reference genes for RT-qPCR studies of gene expression in banana fruit under different experimental conditions. Planta 234: 377–390.
- 39. Obrero A, Die JV, Roman B, Gomez P, Nadal S, et al. (2011) Selection of reference genes for gene expression studies in Zucchini (Cucurbita pepo) using qPCR. J Agric Food Chem 59: 5402–5411.
- 40. Iskandar HM, Simpson RS, Casu RE, Bonnet GD, Maclean DJ, et al. (2004) Comparison of reference genes for quantitative real-time polymerase chain reaction analysis of gene in sugarcane. Plant Mol Biol Rep 22: 325–337.
- 41. Hong SY, Seo PJ, Yang MS, Xiang F, Park CM (2008) Exploring valid reference genes for gene expression studies in Brachypodium distacyon by real-time PCR. BMC Plant Biol 8: 112–122.
- 42. Lee JM, Roche JR, Donaghy DJ, Thrush A, Sathish P (2010) Validation of reference genes for quantitative RT-PCR studies of gene expression in perennial ryegrass (Lolium perenne L.). BMC Mol Biol 11: 8–22.
- 43. Huis R, Hawkins S, Neutelings G (2010) Selection of reference genes for quantitative gene expression normalization in flax (Linum usitatissimum L.). BMC Plant Biol 10: 71–86.
- 44. Yang Y, Hou S, Cui G, Chen S, Wei J, et al. (2010) Characterization of reference genes for quantitative real-time PCR analysis in various tissue of Salvia miltiorrhiza. Mol Biol Rep 37: 507–513.
- 45. Tong Z, Gao Z, Wang F, Zhou J, Zhang Z (2009) Selection of reliable reference genes for gene expression studies in peach using realtime PCR. BMC Mol Biol 10: 71–82.
- 46. McNulty S, Toscano W (1995) Transcriptional regulation of glyceraldehyde-3-phosphate dehydrogenase by 2,3,7,8-tetrachlorodibenzo- p-dioxin. Biochem Biophys Res Commun 212: 165–171.
- 47. Hochstrasser M (2000) Evolution and function of ubiquitin-like protein- conjugation systems. Nature Cell Biology 2: E153–157.
- 48. Stürzenbaum SR, Kille P (2001) Control genes in quantitative molecular biological techniques: the variability of invariance. Comp Biochem Physiol Biochem Mol Biol 130: 281–289.
- 49. B. rapa genome sequence portal. 01: Available: http://www.brassica-rapa.org/BrEMD/microarray_overview.jsp. Accessed 2011 Dec.
- 50. Brassica database. 20: Available: http://brassicadb.org/brad. Accessed 2011 Dec.
- 51. geNorm software manual update. 22: Available: http://medgen.ugent.be/~jvdesomp/genorm/. Accessed 2011 Nov.
- 52. Normfinder software manual updates. 24: Available: http://www.mdl.dk/publicationsnormfinder.htm. Accessed 2011 Nov.