《园艺作物育种学》课程教学资源（学术研究）Validation of reference genes for real-time quantitative PCR normalisation in non-heading Chinese cabbage

CSIRO PUBLISHING Functional Plant Biology,2012,39,342-350 http://dx.doi.org/10.1071/FP11246 Validation of reference genes for real-time quantitative PCR normalisation in non-heading Chinese cabbage Dong XiaoB,Ning-Wen ZhangB,Jian-Jun ZhaoB.C,Guusje BonnemaB.D and Xi-Lin HouD AState Key Laboratory of Crop Genetics and Germplasm Enhancement;Horticultural College, Nanjing Agricultural University,Nanjing,Jiangsu 210095,China. BLaboratory of Plant Breeding,Wageningen University,The Netherlands. CHorticultural College,Hebei Agricultural University,Baoding,Hebei,China. PCorresponding author.Emails:guusje.bonnema@wur.nl;hxl@njau.edu.cn Abstract.Non-heading Chinese cabbage is an important vegetable crop that includes pak choi,caixin and several Japanese vegetables like mizuna,mibuna and komatsuna.Gene expression studies are frequently used to unravel the genetics of complex traits and in such studies the proper selection of reference genes for normalisation is crucial.We assessed the expression of 13 candidate reference genes including ACTIN,ACTIN-1,ACTIN-2,GAPDH,Tub_a,CyP,EFI-a,18SrRNA, UBO,UBC30,PPR,PP24 and MDH.Their expression stabilities were analysed using two programs,geNorm and NormFinder,in 20 different samples that represent four strategic groups.Results showed that no single gene was uniformly expressed in all tested samples.ACTINand CyP are proposed as good reference genes when studying developmental stages. CyP,Tub_a and UBC30 are good reference genes when studying different tissues(from flowering to seed set).CyP and Tub_a are the most stable reference genes under biotic stress treatments using the fungi Peronospora parasitica and Alternaria brassicicola.UBC30,EF/-a and ACTIN are recommended for normalisation in abiotic stress studies,including hormone,salt,drought,cold and heath treatments.Moreover,at least five reference genes(ACTIN,CyP,UBC30,EFI-a and UBO)are required for accurate qRT-PCR data normalisation when studying gene expression across all tested samples. Additional keywords:Brassica rapa ssp.chinensis,gene expression,qRT-PCR,reference genes. Received 29 October 2011,accepted 7 March 2012,published online 24 April 2012 Introduction expression level under all the experimental situations tested The quantification of mRNA (mRNA)transcript levels has (Kim et al.2003;Ding et al.2004;Argyropoulos et al.2006). become an important research tool in recent years.Changes in Use of inappropriate reference genes in relative quantification mRNA transcript levels are crucial during plant developmental of gene expression profiles may lead to erroneous normalisation processes,between different tissues and under changing and consequently,misinterpretation of the results.Therefore,it environmental conditions.Real-time quantitative PCR (qRT- is essential to validate the expression stability of reference genes PCR)has become the most popular method to quantify mRNA in each experimental system transcription levels and to validate whole-genome microarray In plant research, glyceraldehyde-3-phosphate data because of its outstanding accuracy,broad dynamic range dehydrogenase (GAPDH),B-ACTIN (ACTIN),tubulin a and high sensitivity not only in the fields of molecular medicine, (Tub_a)and 18S rRNA were considered to have a constant biotechnology,microbiology and molecular diagnostics but also expression level and as a consequence have been widely used in plant research (Vandesompele et al.2002;Jian et al.2008; as reference genes for normalisation of gRT-PCR data in Paolacci et al.2009).Estimating the expression levels of target various experimental conditions (Kim et al.2003;Ding et al. genes of interest by qRT-PCR depends on endogenous control 2004;Jian et al.2008;Lovdal and Lillo 2009).However,it has genes to normalise qRT-PCR;control genes are also called also been reported that the transcript levels of these genes can reference genes or housekeeping genes (HKGs)(Wierschke change significantly under different experimental conditions et al.2010;Martinez-Beamonte et al.2011).HKGs play a (Czechowski et al.2005;Terrier and Glissant et al.2005; general role in basic cellular processes,such as cell structure Basa et al.2009;Chen et al.2010).Recently,many novel maintenance and primary cellular metabolism and thus,their reference genes have been identified from Affymetrix expression is usually unaffected by external factors.An 'ideal' GeneChip data and Microarray datasets in Arabidopsis.One of reference gene for gRT-PCR has a constant and consistent the findings was that among them F-box protein(F-box),SAND expression level over all samples across different experimental family protein and mitosis protein YLS8 were more stably conditions and different tissues.However,several reports expressed than the commonly used reference genes ACTIN-2, demonstrated that there was no single gene with a constant elongation-factor-1-a (EF/-)and ubiquitin-conjugating Journal compilation CSIRO 2012 www.publish.csiro.au/joumals/fpb

Validation of reference genes for real-time quantitative PCR normalisation in non-heading Chinese cabbage Dong XiaoA,B , Ning-Wen Zhang B , Jian-Jun Zhao B,C , Guusje Bonnema B,D and Xi-Lin HouA,D AState Key Laboratory of Crop Genetics and Germplasm Enhancement; Horticultural College, Nanjing Agricultural University, Nanjing, Jiangsu 210095, China. B Laboratory of Plant Breeding, Wageningen University, The Netherlands. CHorticultural College, Hebei Agricultural University, Baoding, Hebei, China. DCorresponding author. Emails: guusje.bonnema@wur.nl; hxl@njau.edu.cn Abstract. Non-heading Chinese cabbage is an important vegetable cropthatincludes pak choi, caixin and several Japanese vegetables like mizuna, mibuna and komatsuna. Gene expression studies are frequently used to unravel the genetics of complex traits and in such studies the proper selection of reference genes for normalisation is crucial. We assessed the expression of 13 candidate reference genes includingACTIN,ACTIN-1,ACTIN-2,GAPDH, Tub_a,CyP,EF1-a, 18S rRNA, UBQ, UBC30, PPR, PP2A and MDH. Their expression stabilities were analysed using two programs, geNorm and NormFinder, in 20 different samples that represent four strategic groups. Results showed that no single gene was uniformly expressed in all tested samples. ACTIN and CyP are proposed as good reference genes when studying developmental stages. CyP, Tub_a and UBC30 are good reference genes when studying different tissues (from flowering to seed set). CyP and Tub_a are the most stable reference genes under biotic stress treatments using the fungi Peronospora parasitica and Alternaria brassicicola. UBC30, EF1-a and ACTIN are recommended for normalisation in abiotic stress studies, including hormone, salt, drought, cold and heath treatments. Moreover, at least five reference genes (ACTIN,CyP, UBC30, EF1-a and UBQ) are required for accurate qRT–PCR data normalisation when studying gene expression across all tested samples. Additional keywords: Brassica rapa ssp. chinensis, gene expression, qRT-PCR, reference genes. Received 29 October 2011, accepted 7 March 2012, published online 24 April 2012 Introduction The quantification of mRNA (mRNA) transcript levels has become an important research tool in recent years. Changes in mRNA transcript levels are crucial during plant developmental processes, between different tissues and under changing environmental conditions. Real-time quantitative PCR (qRT– PCR) has become the most popular method to quantify mRNA transcription levels and to validate whole-genome microarray data because of its outstanding accuracy, broad dynamic range and high sensitivity not only in the fields of molecular medicine, biotechnology, microbiology and molecular diagnostics but also in plant research (Vandesompele et al. 2002; Jian et al. 2008; Paolacci et al. 2009). Estimating the expression levels of target genes of interest by qRT–PCR depends on endogenous control genes to normalise qRT–PCR; control genes are also called reference genes or housekeeping genes (HKGs) (Wierschke et al. 2010; Martínez-Beamonte et al. 2011). HKGs play a general role in basic cellular processes, such as cell structure maintenance and primary cellular metabolism and thus, their expression is usually unaffected by external factors. An ‘ideal’ reference gene for qRT–PCR has a constant and consistent expression level over all samples across different experimental conditions and different tissues. However, several reports demonstrated that there was no single gene with a constant expression level under all the experimental situations tested (Kim et al. 2003; Ding et al. 2004; Argyropoulos et al. 2006). Use of inappropriate reference genes in relative quantification of gene expression profiles may lead to erroneous normalisation and consequently, misinterpretation of the results. Therefore, it is essential to validate the expression stability of reference genes in each experimental system. In plant research, glyceraldehyde-3-phosphate dehydrogenase (GAPDH), b-ACTIN (ACTIN), tubulin a (Tub_a) and 18S rRNA were considered to have a constant expression level and as a consequence have been widely used as reference genes for normalisation of qRT–PCR data in various experimental conditions (Kim et al. 2003; Ding et al. 2004; Jian et al. 2008; Løvdal and Lillo 2009). However, it has also been reported that the transcript levels of these genes can change significantly under different experimental conditions (Czechowski et al. 2005; Terrier and Glissant et al. 2005; Basa et al. 2009; Chen et al. 2010). Recently, many novel reference genes have been identified from Affymetrix GeneChip data and Microarray datasets in Arabidopsis. One of the findings was that among them F-box protein (F-box), SAND family protein and mitosis protein YLS8 were more stably expressed than the commonly used reference genes ACTIN-2, elongation-factor-1-a (EF1-a) and ubiquitin-conjugating CSIRO PUBLISHING Functional Plant Biology, 2012, 39, 342–350 http://dx.doi.org/10.1071/FP11246 Journal compilation CSIRO 2012 www.publish.csiro.au/journals/fpb

Validation of reference genes in a wide of samples Functional Plant Biology 343 enzyme 10 (UBC10)(Remans et al.2008).In a recent paper a germinated and grown under controlled conditions in pots in a Brassica napus L.microarray database was analysed,which climate room:25C day/20C night temperature,12h light/12h showed that EF/-a.and a new unknown protein I (UP1)were dark cycles.The plants were used to collect tissues under normal the most suitable reference genes among the given set of tissues growth conditions and after biotic and abiotic stress treatments. (Chen et al.2010).Furthermore,two commonly used reference All samples were snapped frozen in liquid nitrogen and kept genes ACTIN-7 and UBC21,plus two new genes,TIP41-like at-80°C until use. protein (TIP41)and PP2A that were selected from a microarray dataset,were identified as the most stable reference gene set for Developmental stages(Ds) normalisation during B.napus embryo maturation (Chen et al. Three young leaves per plant were harvested and leaves of three 2010).A study in Chinese cabbage showed that EF/-o and plants were pooled for each developmental stage:(i)early stage adenine phosphoribosyl-transferase (Apr)were the most stably (third leave present)(Ds1);(ii)before bolting (8 weeks after expressed genes among different tissues (root,stem,heading sowing)(Ds2);and (iii)after bolting (10 weeks after sowing) leaves and lateral sprout)(Qi et al.2010) Ds3). The morphological variation present within Brassica rapa (L.)Hanelt is enormous.This includes the leaves in crops like Different tissues (Dt) heading Chinese cabbages and the leafy types that do not form Eight different tissues including root(Dtl)and stem(Dt2)at the heads (pak choi,caixin and several Japanese vegetables like third leaf stage,leaves after bolting (Dt3,same sample as Ds3) mizuna,mibuna and komatsuna),the enlarged roots of turips, flower buds(Dt4),petiole s(Dt5),stamens(Dt6),pistils(Dt7)and the inflorescences and stems of broccoletto and the seeds of seed pods (Dt8),were collected from three plants and pooled. the oil types.When studying the genetic relationship among accessions using AFLP and SSR marker profiling,clusters or Biotic stress treatments (Bs) groups of accessions were identified that were represented by different crop types,but it was also clear that genetic Two fungi,Peronospora parasitica (P.p)and Alternaria distance was more defined by geographical origin than by crop brassicicola (A.b),were isolated from the leaves of different type (Zhao et al.2007,2010).There is no information about susceptible B.rapa cultivars in the farm of Nanjing selection of reference genes for nommalisation of gRT-PCR Agricultural University,China.Conidial suspensions were results for gene expression studies in Chinese cabbage.With adjusted to Ix 10%sporesmL and Tween-20 was added as a the recently released B.rapa genome sequence (The Brassica surfactant to a final concentration of 0.1%.For each treatment, rapa Genome Sequencing Project Consortium 2011)and 53-week-old seedlings were sprayed either with 50 mL pathogen development of gene expression platforms for B.rapa, suspension,or demi water (as control).Treated seedlings genome-wide large-scale gene expression studies will become were placed in a climate chamber (25C,85%+5%RH, available and will be mined to select reference genes for real 12 hour light/12 hour dark)and the second leaves from three time PCR studies. plants per treatment were harvested and pooled at 48h after In the present study,we selected and validated 13 reference inoculation.The two treatments are referred to as Bsl and Bs2 genes particularly for accurate normalisation ofgRT-PCR results respectively in non-heading Chinese cabbage.These reference genes include seven widely used reference genes in plant research (ACTIN, Abiotic stress treatments (As) ACTIN-2,GAPDH,Tub_a,CyP,EFI-a and 18S rRNA)and six Fifty seedlings (3 weeks old)were sprayed respectively with potential reference genes (ACTIN-1,UBO,UBC30,PPR,PP2A 50 mL solutions,water(as control),SA(2 mmolL,PH=6.5), and MDA)that were identified based on their stability in ABA(50 umol),NaCl(200 mM),H2O2(100 uM)and Mannitol expression studies comparing plant developmental stages, (400 mM).For salt (Asl)and drought (As2)stress treatments, different tissues or various environmental stimuli including leaves were harvested at 12 h(for NaCl and Mannitol)after stress biotic and abiotic stress (Brunner et al.2004;Czechowski treatments.For hormone treatments.leaves were harvested at et al.2005;Reid et al.2006).The 13 genes were tested in 20 6h after SA treatment(As3),24h after ABA treatment(As4) different samples,including three developmental stages,eight and 24 h after H2O2 treatment(As5).In addition,leaves from different tissues harvested at the mature plant developmental 50 seedlings(3 weeks old)that were exposed to cold(4C)and stage (between flowering and seed set)and from seedlings heat shock (40C)were harvested 2h after temperature stress exposed to two biotic stresses and seven abiotic stress treatments (As6 and As7). treatments,ranging from hormone-,salt-,drought-,till temperature stress treatments.We used the statistical algorithms RNA isolation,quality control and cDNA synthesis geNorm (Vandesompele et al.2002)and NormFinder (Andersen Total RNA was isolated by RNA simple Total RNA Kit et al.2004),which have been widely employed to select the extractions (Bio Teke.Beijing.China).Genomic DNA best suitable reference genes from given biological samples contaminations were effectively removed using RNase-free (Silver et al.2006;Wierschke et al.2010). DNase I treatment(Invitrogen,Carlsbad,CA,USA)according to manufacturer's instructions,as melting curve gave single peak and genomic amplification was larger than RNA/cDNA Materials and methods amplification(see Fig.S1,available as Supplementary Material Seeds of one pakchoi inbred line (Brassica rapa ssp.chinensis to this paper).RNA integrity was electrophoretically verified (L.)Hanelt;Suzhou Qing,a non-heading Chinese cabbage)were by agarose gel and by 260/280nm absorption ratio 1.9~2.1

enzyme 10 (UBC10) (Remans et al. 2008). In a recent paper a Brassica napus L. microarray database was analysed, which showed that EF1-a and a new unknown protein 1 (UP1) were the most suitable reference genes among the given set of tissues (Chen et al. 2010). Furthermore, two commonly used reference genes ACTIN-7 and UBC21, plus two new genes, TIP41-like protein (TIP41) and PP2A that were selected from a microarray dataset, were identified as the most stable reference gene set for normalisation during B. napus embryo maturation (Chen et al. 2010). A study in Chinese cabbage showed that EF1-a and adenine phosphoribosyl-transferase (Apr) were the most stably expressed genes among different tissues (root, stem, heading leaves and lateral sprout) (Qi et al. 2010). The morphological variation present within Brassica rapa (L.) Hanelt is enormous. This includes the leaves in crops like heading Chinese cabbages and the leafy types that do not form heads (pak choi, caixin and several Japanese vegetables like mizuna, mibuna and komatsuna), the enlarged roots of turnips, the inflorescences and stems of broccoletto and the seeds of the oil types. When studying the genetic relationship among accessions using AFLP and SSR marker profiling, clusters or groups of accessions were identified that were represented by different crop types, but it was also clear that genetic distance was more defined by geographical origin than by crop type (Zhao et al. 2007, 2010). There is no information about selection of reference genes for normalisation of qRT–PCR results for gene expression studies in Chinese cabbage. With the recently released B. rapa genome sequence (The Brassica rapa Genome Sequencing Project Consortium 2011) and development of gene expression platforms for B. rapa, genome-wide large-scale gene expression studies will become available and will be mined to select reference genes for real time PCR studies. In the present study, we selected and validated 13 reference genes particularly for accurate normalisation of qRT–PCR results in non-heading Chinese cabbage. These reference genes include seven widely used reference genes in plant research (ACTIN, ACTIN-2, GAPDH, Tub_a, CyP, EF1-a and 18S rRNA) and six potential reference genes (ACTIN-1, UBQ, UBC30, PPR, PP2A and MDH) that were identified based on their stability in expression studies comparing plant developmental stages, different tissues or various environmental stimuli including biotic and abiotic stress (Brunner et al. 2004; Czechowski et al. 2005; Reid et al. 2006). The 13 genes were tested in 20 different samples, including three developmental stages, eight different tissues harvested at the mature plant developmental stage (between flowering and seed set) and from seedlings exposed to two biotic stresses and seven abiotic stress treatments, ranging from hormone-, salt-, drought-, till temperature stress treatments. We used the statistical algorithms geNorm (Vandesompele et al. 2002) and NormFinder (Andersen et al. 2004), which have been widely employed to select the best suitable reference genes from given biological samples (Silver et al. 2006; Wierschke et al. 2010). Materials and methods Seeds of one pakchoi inbred line (Brassica rapa ssp. chinensis (L.) Hanelt; Suzhou Qing, a non-heading Chinese cabbage) were germinated and grown under controlled conditions in pots in a climate room: 25
C day/20
C night temperature, 12 h light/12 h dark cycles. The plants were used to collect tissues under normal growth conditions and after biotic and abiotic stress treatments. All samples were snapped frozen in liquid nitrogen and kept at 80
C until use. Developmental stages (Ds) Three young leaves per plant were harvested and leaves of three plants were pooled for each developmental stage: (i) early stage (third leave present) (Ds1); (ii) before bolting (8 weeks after sowing) (Ds2); and (iii) after bolting (10 weeks after sowing) (Ds3). Different tissues (Dt) Eight different tissues including root (Dt1) and stem (Dt2) at the third leaf stage, leaves after bolting (Dt3, same sample as Ds3), flower buds (Dt4), petiole s(Dt5), stamens (Dt6), pistils (Dt7) and seed pods (Dt8), were collected from three plants and pooled. Biotic stress treatments (Bs) Two fungi, Peronospora parasitica (P.p) and Alternaria brassicicola (A.b), were isolated from the leaves of different susceptible B. rapa cultivars in the farm of Nanjing Agricultural University, China. Conidial suspensions were adjusted to 1  105 spores mL–1 and Tween-20 was added as a surfactant to a final concentration of 0.1%. For each treatment, 53-week-old seedlings were sprayed either with 50 mL pathogen suspension, or demi water (as control). Treated seedlings were placed in a climate chamber (25
C, 85%  5% RH, 12 hour light/12 hour dark) and the second leaves from three plants per treatment were harvested and pooled at 48 h after inoculation. The two treatments are referred to as Bs1 and Bs2 respectively. Abiotic stress treatments (As) Fifty seedlings (3 weeks old) were sprayed respectively with 50 mL solutions, water (as control), SA (2 mmol L1 , PH = 6.5), ABA (50 mmol), NaCl (200 mM), H2O2 (100 mM) and Mannitol (400 mM). For salt (As1) and drought (As2) stress treatments, leaves were harvested at 12 h (for NaCl and Mannitol) after stress treatments. For hormone treatments, leaves were harvested at 6 h after SA treatment (As3), 24 h after ABA treatment (As4) and 24 h after H2O2 treatment (As5). In addition, leaves from 50 seedlings (3 weeks old) that were exposed to cold (4
C) and heat shock (40
C) were harvested 2 h after temperature stress treatments (As6 and As7). RNA isolation, quality control and cDNA synthesis Total RNA was isolated by RNA simple Total RNA Kit extractions (Bio Teke, Beijing, China). Genomic DNA contaminations were effectively removed using RNase-free DNase I treatment (Invitrogen, Carlsbad, CA, USA) according to manufacturer’s instructions, as melting curve gave single peak and genomic amplification was larger than RNA/cDNA amplification (see Fig. S1, available as Supplementary Material to this paper). RNA integrity was electrophoretically verified by agarose gel and by 260/280 nm absorption ratio 1.9~2.1 Validation of reference genes in a wide of samples Functional Plant Biology 343

344 Functional Plant Biology D.Xiao et al. (see Fig.S2).The first-strand ofcDNA was synthesised by reverse stress treatments)and were also treated as one group for transcribing I ug of total RNA in a final reaction volume of analysis of candidate reference gene stabilities.Data or mean 20 uL using the M-MLV reverse transcriptase (Takara,Dalian, Ct values obtained from gRT-PCR were transformed to China)according to the manufacturer's protocol and diluted quantities with PCR efficiency derived straight from 1:10 before use in gRT-PCR assays.The concentration and amplification plots using LinReg (ver.7.0;Amsterdam,The quality ofeach RNA and cDNA sample was also measured using Netherlands)software (Ramakers et al.2003).The normalised the nucleic acid analytic apparatus K6000 (Bio Photometer, data were imported and analysed by two stability analysis Eppendorf,Germany). programs for reference genes,geNorm ver.3.4 (Vandesompele et al.2002)and NormFinder (Andersen et al.2004)for ranking Primer design and gRT-PCR the reference genes Specific primer pairs were designed for 10 single genes(GAPDH, geNorm determines the gene stability measure (M)value Tub_d,Cyclophilin (CyP),EFI-a,18S rRNA,poly ubiquitin for each gene,based on the average pair-wise variation for a enzyme(UBO),UBC30,Pentatricopeptide repeat(PPR),PP2A particular gene with all the other tested genes.Thus,genes can be and Malate dehydrogenase(MDH))and three orthologous genes ranked according to their expression stability through the (ACTIN,ACTIN-1 and ACTIN-2)of the ACTIN gene family.The stepwise exclusion of the least stable gene.The genes with an sequences of the genes were retrieved from Arabidopsis thaliana M value were arbitrarily suggested to be lower than 1.5;genes (L.)Heynh.and blasted against EST(expressed sequence tag) with the lowest M values have the most stable expression.A pair- libraries (Brassica)via nucleotide blast.The Arabidopsis wise stability measure aims at determining the benefit of adding consensus sequences were used for comparison with B.rapa extra reference genes for the normalisation process.For this,an EST sequences,if available,to reveal the exon-intron structure. arbitrary cut off value of0.15 for pair-wise variation (Vn/Vn+1) When no B.rapa L.ssp.chinensis (L.)Hanelt EST sequence of normalisation factor (NF)(NFn and NFn+1)is calculated, was available the gene structure was obtained by comparison of reflecting the effect of including additional(n+1)genes. the A.thaliana sequence with Brassica napus L.and Brassica The second different statistical algorithm software, oleracea L.EST sequences assuming that the exon-exon NormFinder,generates a stability value for each gene,which boundaries are conserved between A.thaliana and B.napus/ is a direct measure for its estimated expression variation.It ranks B.rapa/B.oleracea (see Fig.S3).Sequence comparisons were the stability level of each candidate gene and highlights the most done by DNASTAR,Lasergene 9.1 (Lasergene,Madison,WI stably expressed gene with the lowest stability value by using a USA)and primers were designed using Primer Express 2.0 model-based variance estimation approach. software (PE Applied Biosystems,Foster,CA,USA)under default parameters.Primer sequences and exon-exon junctions Results of the reference genes are listed in Table 1.Primer specificity and Expression profiling of the candidate genes DNA contamination were visualised by separating PCR products from cDNA and DNA on agarose gel (Fig.S1A)and when only a Melting curve analysis of the amplification products confirmed single band was observed,the band was purified to be template for that the primers amplified a single PCR product(Fig.S3).The preparing the standard curves(which takes into account primer standard curves for each of the candidate reference genes were efficiency)using gRT-PCR (Ramakers et al.2003).The PCR found to have R2>0.995 (Table 1),indicating a strong reaction efficiencies (E)were calculated using the equation linear relationship between the detected Ct values and the E=10(-1/slope)(Lekanne Deprez et al.2002).This calculation corresponding relative amount of cDNA in all the PCR method results in efficiencies ranging from 95 to 108%.The reactions.Based on the slopes of the standard curves,the 13 standard curve was generated using a dilution series of the Dsl gene assays were found to have PCR efficiencies >90.7% (leaf developmental stage 1)sample over at least five dilution (Table 1).It was apparent that each candidate reference gene points (Fig.S4).gRT-PCR was performed in triplicate on a had variable Ct values in the wide variety of samples tested,as the 32-position carousel (Light Cycler)with the Light Cycler-RNA Ct values ranged widely from 14.56(18S rRNA)to 39.53(PP24) amplification kit SYBR Green I(Roche,Mannheim,Germany) across all samples (Table 2).None of the 13 candidate reference and conducted in 25 uL reaction volumes containing 30ng uL genes had a uniform expression over all samples tested.As a cDNA sample,along with an RNA template control in parallel consequence,for accurate normalisation of gene expression in for each gene.The thermal cycling consisted of 95C for 2 min different experimental conditions,specific sets ofreference genes and 40 cycles of95C for 20s,55C for 20s and 72C for 20s.After are needed. the PCR a melting curve was generated to check the specificity of the amplified fragment.Data analysis was performed with geNorm analysis the Rotor-gene 6 ver.6.1 software(Applied Biosystems).All the The geNorm program conducts sequential elimination ofthe least cycle threshold(Ct)values from one gene were determined at stable gene in any given experimental group,resulting in the the same threshold fluorescence value of 0.2.The single-peak exclusion of all but the two most stable genes in each strategic melting curves obtained using the 13 primer pairs to amplify the group(Table 3).For all 20 samples tested,ACTIN and CyP were candidate reference genes are displayed in Fig.S1B. the most stable genes,followed by UBC30,EFI-a and UBO. In contrast,PP2 A was the least stable gene tested.The two most Data analyses stably expressed genes when comparing samples of different The 20 samples were divided into four strategic groups developmental stages were ACTIN and CyP that were also most (developmental stages,different tissues,biotic and abiotic stable when all 20 samples were analysed together,followed by

(see Fig. S2). Thefirst-strand of cDNA was synthesised by reverse transcribing 1 mg of total RNA in a final reaction volume of 20 mL using the M-MLV reverse transcriptase (Takara, Dalian, China) according to the manufacturer’s protocol and diluted 1 : 10 before use in qRT–PCR assays. The concentration and quality of each RNA and cDNA sample was also measured using the nucleic acid analytic apparatus K6000 (Bio Photometer, Eppendorf, Germany). Primer design and qRT–PCR Specific primer pairs were designed for 10 single genes (GAPDH, Tub_a, Cyclophilin (CyP), EF1-a, 18S rRNA, poly ubiquitin enzyme (UBQ), UBC30, Pentatricopeptide repeat (PPR), PP2A and Malate dehydrogenase (MDH)) and three orthologous genes (ACTIN, ACTIN-1 and ACTIN-2) of the ACTIN gene family. The sequences of the genes were retrieved from Arabidopsis thaliana (L.) Heynh. and blasted against EST (expressed sequence tag) libraries (Brassica) via nucleotide blast. The Arabidopsis consensus sequences were used for comparison with B. rapa EST sequences, if available, to reveal the exon-intron structure. When no B. rapa L. ssp. chinensis (L.) Hanelt EST sequence was available the gene structure was obtained by comparison of the A. thaliana sequence with Brassica napus L. and Brassica oleracea L. EST sequences assuming that the exon-exon boundaries are conserved between A. thaliana and B. napus/ B. rapa/B. oleracea (see Fig. S3). Sequence comparisons were done by DNASTAR, Lasergene 9.1 (Lasergene, Madison, WI, USA) and primers were designed using Primer Express 2.0 software (PE Applied Biosystems, Foster, CA, USA) under default parameters. Primer sequences and exon-exon junctions of the reference genes are listed in Table 1. Primer specificity and DNA contamination were visualised by separating PCR products from cDNA and DNA on agarose gel (Fig. S1A) and when only a single band was observed, the band was purified to be template for preparing the standard curves (which takes into account primer efficiency) using qRT–PCR (Ramakers et al. 2003). The PCR reaction efficiencies (E) were calculated using the equation E = 10(–1/slope) (Lekanne Deprez et al. 2002). This calculation method results in efficiencies ranging from 95 to 108%. The standard curve was generated using a dilution series of the Ds1 (leaf developmental stage 1) sample over at least five dilution points (Fig. S4). qRT–PCR was performed in triplicate on a 32-position carousel (Light Cycler) with the Light Cycler-RNA amplification kit SYBR Green I (Roche, Mannheim, Germany) and conducted in 25 mL reaction volumes containing 30ng mL–1 cDNA sample, along with an RNA template control in parallel for each gene. The thermal cycling consisted of 95
C for 2 min and 40 cycles of 95
C for 20s, 55
C for 20s and 72
C for 20s. After the PCR a melting curve was generated to check the specificity of the amplified fragment. Data analysis was performed with the Rotor-gene 6 ver. 6.1 software (Applied Biosystems). All the cycle threshold (Ct) values from one gene were determined at the same threshold fluorescence value of 0.2. The single-peak melting curves obtained using the 13 primer pairs to amplify the candidate reference genes are displayed in Fig. S1B. Data analyses The 20 samples were divided into four strategic groups (developmental stages, different tissues, biotic and abiotic stress treatments) and were also treated as one group for analysis of candidate reference gene stabilities. Data or mean Ct values obtained from qRT–PCR were transformed to quantities with PCR efficiency derived straight from amplification plots using LinReg (ver. 7.0; Amsterdam, The Netherlands) software (Ramakers et al. 2003). The normalised data were imported and analysed by two stability analysis programs for reference genes, geNorm ver. 3.4 (Vandesompele et al. 2002) and NormFinder (Andersen et al. 2004) for ranking the reference genes. geNorm determines the gene stability measure (M) value for each gene, based on the average pair-wise variation for a particular gene with all the other tested genes. Thus, genes can be ranked according to their expression stability through the stepwise exclusion of the least stable gene. The genes with an M value were arbitrarily suggested to be lower than 1.5; genes with the lowest M values have the most stable expression. A pair￾wise stability measure aims at determining the benefit of adding extra reference genes for the normalisation process. For this, an arbitrary cut off value of 0.15 for pair-wise variation (Vn/Vn + 1) of normalisation factor (NF) (NFn and NFn + 1) is calculated, reflecting the effect of including additional (n + 1) genes. The second different statistical algorithm software, NormFinder, generates a stability value for each gene, which is a direct measure for its estimated expression variation. It ranks the stability level of each candidate gene and highlights the most stably expressed gene with the lowest stability value by using a model-based variance estimation approach. Results Expression profiling of the candidate genes Melting curve analysis of the amplification products confirmed that the primers amplified a single PCR product (Fig. S3). The standard curves for each of the candidate reference genes were found to have R2 0.995 (Table 1), indicating a strong linear relationship between the detected Ct values and the corresponding relative amount of cDNA in all the PCR reactions. Based on the slopes of the standard curves, the 13 gene assays were found to have PCR efficiencies 90.7% (Table 1). It was apparent that each candidate reference gene had variable Ct values in the wide variety of samples tested, as the Ct values ranged widely from 14.56 (18S rRNA) to 39.53 (PP2A) across all samples (Table 2). None of the 13 candidate reference genes had a uniform expression over all samples tested. As a consequence, for accurate normalisation of gene expression in different experimental conditions, specific sets of reference genes are needed. geNorm analysis The geNorm program conducts sequential elimination of the least stable gene in any given experimental group, resulting in the exclusion of all but the two most stable genes in each strategic group (Table 3). For all 20 samples tested, ACTIN and CyP were the most stable genes, followed by UBC30, EF1-a and UBQ. In contrast, PP2A was the least stable gene tested. The two most stably expressed genes when comparing samples of different developmental stages were ACTIN and CyP that were also most stable when all 20 samples were analysed together, followed by 344 Functional Plant Biology D. Xiao et al

Validation of reference genes in a wide of samples Functional Plant Biology 345 24.201 001m01 06101 06001 景 00.201 屋 层 层 季 层 06.101 08- 617'S- 0845 666.0 866.0 166:0 多 多 666.0 多 多 多 富 S66:0 (dq)azis 监 兰 公 登 图 壁 的 母 监 宝 壁 8 亨 寻 8 8 艺 8 8 三 君 三 25-c2 罩 uonoun t uoxa guox 8 a 8 8 的 的的 A6 9 A 多 8 198 86 8 (asI3AaI pue 10/1.i/t//./en ODOOVOLLLVDVDDVDDV 0yem10010.1.216m I'ELStbOdV 1:1620608 I'8I0SSW I'8t186EdV T669180 -092410 L6604S0:1 17606c00 18681760V 10641Oy 8866170103 89517613 56690110 ywoy snool onooyuo s souo1 u pauuups 'uLSV 3usn uanbas apnoopnu punyop'y woy poumqo sem neA- HddV9 子 100XM VN S8I E HOW

Table 1. Primers and PCR efficiency for the 13 selected candidate reference genes SymbolA Arabidopsis orthologue locus Accession number Primer sequence 50–30 (forward and reverse) Tm (
C) Junction BLASTnB Amplicon size (bp) R2 Slope PCR efficiency (%) ACTIN AT5G09810 AF111812 GGAGCTGAGAGATTCCGTTG 60 Exon4 0.0 158 0.999 –3.410 96.50 GAACCACCACTGAGGACGAT 60 ACTIN-1 AT2G37620 AF044573.1 CCAACAGAGAGAAGATGACCC 59 Exon3 e-149 95 0.998 –3.150 107.70 ACTGGCGTAAAGGGAGAGG 59 ACTIN-2 AT3G18780 BG732274.1 ATCGAGCATGGTGTTGTGAG 60 Exon2 e-140 132 0.997 –3.219 104.00 GGCCTTTGGGTTAAGAGGAG 60 GAPDH AT1G12900 AB331373 TCCACCATTGATTCTTCTCTG 58 Exon5 0.0 108 0.999 –3.293 101.20 TCAGCCAAATCAACAACTCTC 58 CyP AT2G16600 M55018.1 AGGAGGAGATTTCACCGC 58 Exon1 0.0 232 0.999 –3.312 100.40 TCTCTAACGACATCCATCCC 58 EF1-a AT5G60390 AF398148.1 TCTGGAAAAGAGATTGAGAAGG 59 Exon3 e-107 129 0.999 –3.465 94.40 AACAGCGAAACGACCCAAT 61 Tub _ a AT5G19780 AC189186.2 TTTGGGTTCTCTCTTGCTAG 59 Exon3 Exon 4 0.0 143 0.999 –3.267 102.30 CGAGTAGAGAATGAGTTGAG 59 18S rRNA AT3G41768 AF513990.1 ATTGACGGAAGGGCACCAC 60 Exon1 0.0 158 0.999 –3.568 90.70 TCGCTCCACCAACTAAGAAC 59 UBC30 AT5G56150 U17250.1 TGAAAGAGCAGTGGAGCC 58 Exon4 Exon5 e-117 122 0.999 –3.480 93.80 GGTCTGTCTTGTAGGTGTGAGC 59 UBQ AT2G36170 L21898.1 CAGCCAAGGTACGACCATCT 60 Exon4 Exon5 e-149 165 0.999 –3.380 97.60 TATTCGTGAAGACGCTGACG 60 PPR AT1G62860 FJ455099.1 AAGAGGGTAGATGATGGAATG 56 Exon2 e-151 180 0.999 –3.333 99.50 TTACAAGTGACGACATTAGGG 55 PP2A AT1G69960 AC240932.1 AGGCTACACGTTCGGACAAG 60 Exon5 2e-152 142 0.997 –3.431 95.60 TGGGGCACTAAACACAGTCA 60 MDH AT1G62480 CB331882 CGAGATGACACCACCAAAGAC 61 Exon2 4e-10 157 0.995 –3.278 101.90 GGTTTCATCTGCTTCTTCGG 60 AAll the selected reference genes were named according to the most similar orthologue locus from A. thaliana. BE-value was obtained from the A. thaliana nucleotide sequence using BLASTn. Validation of reference genes in a wide of samples Functional Plant Biology 345

346 Functional Plant Biology D.Xiao et al. 00FS981 CH016191 00FS981 610+6091 PN019491 17.0116.61 770165.61 6104252 6r019652 6S0769#2 22012102 S#:0HS812 11:0168:62 -0115:52 KE.012592 8#.011592 ￥。：049512 20-.92 H.018212 8:018192 90.0161 0.1112 80:01299t 110+3LL5 5Z.0+8953 喜 6001110.82 22:018982 210422 2101.0M 6E'OF LE'IE #4:010000 00192.0 D0升260 800F E8'8 204101 00t 000190 8P0T13.2E 800 19 20102 #80122.23 52042928 #0US.9 #2:0TG2*2 90013522 t104S.52 1-01622 90:01692 0014902 04:0132 Z:022252 5:0165.22 90.04122 0.0129 0:0125:12 33.0T1522 #8:018202 8001.55.92 60429 650199 65:0139 2019 11018 250t 890T16.t2 150100 SI0 00T13 ST-3+99.53 91'0F9S'I 08.0+59.61 Z.:01.61 08.0199.61 60.012651 .:011.31 01:01561 10.0116.51 6E.031 :012602 H01201 66:012661 61:01*402 .011102 2.0-9202 IS OFtLOC 210121-12 LE'I S8'S 200151-08 250106.52 2025.9 900121-08 860100.2 03 8'OF SI'S 10: 92.01.100.62 20562 H1:010002 101612 29-014822 -0+1612 00.0401.61 91'1FES'I 21-010022 00100.31 2.02961 2205 69.0a5112 101#9212 51-019.22 27092 910+C122 8Z.0512 P'SFLO SP01161 I00180.81 K10110.01 080194 20138 H1OT19.81 C001259i 220T1. 2015.51 HH18.81 101012 620T18.81 01-0125.:02 2P0158.61 #:01191.51 60.015-02 51:012E81 02:016502 10.019000 1汉：016502 H:0195.81 22.0122 002212 6001222 E.0T001 8'0F8I'S 6201602 ?+寸x个 66.101261 X-0160.92 #Z0t5852 YE04168Z 20010 22096212 830416.52 20102 01010 5-018.2 200T02 0115 C10+1192 62042562 820125:62 270H3252 S10H1962 S0011100 S004T900 0016252 CC01182 00.23 1t0452c0 00T11102 SEUT博CTE 18-010662 P'SFLO 10014102 IE0119.61 61-0102:2 61011261 I0F961 0210105.01 ZI'OFSE81 0066sl 8TOFS181 800F8081 70116.2 1E0189212 EEOFEEEZ 91010.02 21016602 CH018612 ZN010002 duns sdnora oons

Table 2. Ct values (
s.d.) of the 20 analysed samples ordered according to their respective strategic groups Strategic groups Sample ACTIN ACTIN-1 ACTIN-2 GAPDH Cyp EF1-a Tub_a 18S rRNA UBC30 UBQ PPR PP2A MDH codes CT ± s.d. CT ± s.d. CT ± s.d. CT ± s.d. CT ± s.d. CT ± s.d. CT ± s.d. CT ± s.d. CT ± s.d. CT ± s.d. CT ± s.d. CT ± s.d. CT ± s.d. Developmental Ds1 20.17 ± 0.01 32.58 ± 0.25 28.09 ± 0.13 20.15 ± 0.09 19.1 ± 0.36 21.45 ± 0.14 27.12 ± 0.10 25.58 ± 0.16 26.77 ± 0.08 24.29 ± 0.24 28.64 ± 0.07 27.87 ± 0.13 22.64 ± 0.14 stages (Ds) Ds2 18.64 ± 0.17 29.41 ± 0.15 25.35 ± 0.24 18.32 ± 0.19 18.08 ± 0.01 20.00 ± 0.14 25.85 ± 1.37 14.56 ± 0.16 25.66 ± 0.39 22.53 ± 0.06 28.68 ± 0.22 25.74 ± 0.1 19.08 ± 0.05 Ds3 19.64 ± 0.31 30.81 ± 0.05 25.91 ± 0.38 20.49 ± 0.20 18.81 ± 0.14 21.91 ± 0.1 30.13 ± 0.06 19.68 ± 0.80 25.68 ± 0.79 25.5 ± 0.13 31.37 ± 0.39 26.38 ± 0.37 18.65 ± 0.04 Different tissues Dt1 23.20 ± 0.19 33.08 ± 0.13 28.59 ± 0.02 31.88 ± 0.41 20.26 ± 0.40 23.25 ± 0.34 27.40 ± 0.57 20.55 ± 0.04 26.46 ± 0.5 24.49 ± 0.11 32.25 ± 0.12 39.53 ± 0.53 24.96 ± 0.13 (between flowering Dt2 19.27 ± 0.19 29.87 ± 0.74 24.96 ± 0.22 20.06 ± 0.01 18.79 ± 0.32 22.55 ± 0.42 26.52 ± 0.24 19.34 ± 0.23 24.38 ± 0.13 24.69 ± 0.26 30.37 ± 0.16 26.34 ± 0.48 16.14 ± 0.16 and seed set) (Dt) Dt3 19.64 ± 0.31 30.81 ± 0.05 25.91 ± 0.38 20.49 ± 0.20 18.81 ± 0.14 21.91 ± 0.1 30.13 ± 0.06 19.68 ± 0.80 25.68 ± 0.79 25.5 ± 0.13 31.37 ± 0.39 26.38 ± 0.37 18.65 ± 0.04 Dt4 17.53 ± 0.10 27.29 ± 0.40 23.08 ± 0.39 18.76 ± 0.34 16.52 ± 0.03 19.10 ± 0.00 24.10 ± 0.48 16.92 ± 0.09 22.65 ± 1.09 20.64 ± 0.07 30.00 ± 0.44 23.56 ± 0.44 14.85 ± 0.03 Dt5 18.35 ± 0.12 26.57 ± 0.16 25.20 ± 0.25 19.83 ± 0.05 17.14 ± 0.26 21.53 ± 1.16 25.12 ± 0.77 18.31 ± 0.40 23.13 ± 0.25 23.91 ± 0.26 30.26 ± 0.01 26.1 ± 0.21 16.79 ± 0.19 Dt6 18.99 ± 0.09 28.44 ± 0.22 22.50 ± 0.10 24.42 ± 0.23 18.09 ± 0.78 22.00 ± 0.12 24.06 ± 0.41 19.47 ± 0.10 22.16 ± 0.42 24.28 ± 0.24 30.92 ± 0.39 25.38 ± 0.49 16.56 ± 0.34 Dt7 18.15 ± 0.18 29.68 ± 0.13 23.54 ± 0.17 20.47 ± 0.03 17.97 ± 0.22 18.80 ± 0.07 24.06 ± 0.56 17.94 ± 0.07 22.88 ± 0.11 20.94 ± 0.22 28.83 ± 0.08 24.28 ± 0.34 17.74 ± 0.23 Dt8 18.08 ± 0.08 28.05 ± 0.20 24.36 ± 0.02 21.26 ± 0.09 17.78 ± 0.23 19.62 ± 0.25 25.15 ± 0.28 18.44 ± 0.39 23.8 ± 0.84 21.59 ± 0.38 31.08 ± 0.26 26.18 ± 0.60 14.79 ± 0.22 Biotic stress (Bs) Bs1 23.91 ± 0.27 32.59 ± 0.01 29.25 ± 0.79 23.73 ± 0.29 21.45 ± 0.28 23.34 ± 0.22 27.27 ± 0.26 20.92 ± 0.56 27.58 ± 0.41 24.54 ± 0.47 30.13 ± 0.01 37.19 ± 0.06 27.2 ± 0.19 Bs2 21.65 ± 0.55 32.79 ± 0.43 26.74 ± 0.47 23.22 ± 0.09 19.88 ± 0.06 22.58 ± 0.55 25.51 ± 0.77 18.43 ± 0.14 25.15 ± 0.57 23.17 ± 0.09 28.76 ± 0.30 34.47 ± 0.11 22.9 ± 0.16 Abiotic stress (As) As1 21.63 ± 0.31 28.13 ± 0.04 26.13 ± 0.27 21.40 ± 0.31 18.42 ± 0.52 21.34 ± 0.31 25.04 ± 0.02 19.75 ± 0.11 24.43 ± 0.34 22.49 ± 0.73 32.71 ± 0.48 34.21 ± 1.75 23.57 ± 0.34 As2 20.85 ± 0.31 28.69 ± 0.36 26.11 ± 0.16 25.18 ± 0.28 18.81 ± 0.44 21.17 ± 0.39 25.77 ± 0.84 19.92 ± 0.59 23.91 ± 0.68 23.21 ± 0.06 32.61 ± 0.08 34.38 ± 0.23 25.96 ± 0.49 As3 23.33 ± 0.33 32.25 ± 0.31 29.72 ± 0.29 26.39 ± 0.29 21.44 ± 0.13 24.64 ± 0.13 25.43 ± 0.36 20.34 ± 0.19 27.67 ± 0.64 24.54 ± 0.34 29.39 ± 0.42 36.62 ± 0.03 24.69 ± 0.59 As4 20.38 ± 0.16 28.11 ± 0.05 25.88 ± 0.73 23.18 ± 0.10 18.81 ± 0.29 22.65 ± 0.19 25.56 ± 0.31 20.11 ± 0.34 24.71 ± 0.33 23.62 ± 0.83 32.18 ± 0.34 34.38 ± 0.46 23.32 ± 0.65 As5 20.49 ± 0.42 28.33 ± 0.02 26.59 ± 0.26 23.54 ± 0.22 20.53 ± 0.10 22.79 ± 0.22 26.07 ± 0.65 20.26 ± 0.32 25.05 ± 0.50 24.73 ± 0.70 32.22 ± 0.54 34.35 ± 1.29 22.56 ± 0.5 As6 21.98 ± 0.43 31.34 ± 0.35 27.73 ± 0.28 22.84 ± 0.29 19.83 ± 0.32 22.12 ± 0.16 27.00 ± 0.26 20.73 ± 0.51 25.34 ± 0.15 22.51 ± 0.23 32.62 ± 0.37 37.75 ± 0.04 23.12 ± 0.22 As7 20.00 ± 0.42 27.90 ± 0.31 25.23 ± 0.22 19.32 ± 0.39 17.68 ± 0.34 21.75 ± 0.28 23.95 ± 0.23 19.47 ± 0.38 24.81 ± 0.09 23.28 ± 0.84 30.45 ± 0.04 35.68 ± 0.25 24.85 ± 0.49 346 Functional Plant Biology D. Xiao et al

Validation of reference genes in a wide of samples Functional Plant Biology 347 Table 3. Expression stability values of reference genes ranked by geNorm and NormFinder for the four strategic groups and all 20 samples together (ranking in parentheses) Gene geNorm NormFinder Total Developmental Different Biotic Abiotic Total Developmental Different Biotic Abiotic stages tissues stress stress stages tissues stress stress ACTIN 0.61(1) 0.12(1) 0.56(4) 0.01(1) 0.46(2) 0.09(1) 0.20(3) 0.17(1) 0.05(5 0.26(3) ACTIN-I 1.21(8) 0.36(5) 0.52(3) 0.71(12) 0.51(3) 0.90(9) 0.34(7 0.463) 0.74(13) 0.32(5) ACTIN-2 1.11(7 0.53(7 0.82(7 0.43(8) 0.65(7 0.85(8) 0.33(6 0.56(6 0.40(7) 0.57(10) GAPDH 1.02(6) 0.20(2) 1.03(9) 0.31(7) 0.70(8) 0.68(5) 0.29(5) 0.69(9) 0.56(10) 0.50(8) CyP 0.61(1) 0.12(1) 0.45(2) 0.12(5 0.54(4) 0.50(4) 0.05(1) 0.41(2) 0.02(3) 0.32(6 EFI-a 0.74(3) 0.23(3) 0.66(5) 0.22(6) 0.281) 0.39(2) 0.28(4) 0.51(4) 0.40(8) 0.32(4) Tub a 0.92(5) 0.77(9) 0.31(1) 0.01(1) 0.60(6 0.85(7) 0.87(11) 0.53(5) 0.04(4) 0.34(7) 18S rRNA 1.43(10) 1.54(12) 1.22(11) 0.58(10) 0.57(5) 1.17(12) 3.44(13) 1.03(12) 0.62(11) 0.18(1) UBC30 0.72(2) 0.34(4) 0.31(1) 0.061(2) 0.28(1) 0.46(3) 0.52(8) 0.62(8) 0.00(1) 0.21(2) UBO 0.86(4) 0.66(8) 0.91(8) 0.07(3) 0.74(9) 0.79(6 0.59(9) 0.81(10) 0.00(2) 0.54(9) PPR 1.31(9) 0.84(10) 0.77(6 0.08(4) 0.95(12) 1.05(10 0.97(12) 0.60(7) 0.11(6) 0.91(13) PP2A 1.90(12) 0.44(6) 1.54(12) 0.50(9) 0.80(10) 2.51(13) 0.05(2) 2.21(13) 0.47(9) 0.68(11) MDH 1.56(11) 0.92(11) 1.12(10) 0.65(11) 0.86(11) 1.16(11) 0.68(10) 0.83(11) 0.73(12) 0.68(12) GAPDH,EFI-o.and UBC30.Furthermore,the genes ACTIN. for accurate normalisation,i.e.ACTIN,CyP,UBC30,EFI-o and CyP and GAPDH with lowest M values displayed less variation UBO.When comparing developmental stages,ACTIN and Cyp (from 0.I to 0.3)than M values ofother genes.UBC30 and Tub_a formed the optimal set of reference genes,whereas in different were the most stable genes when gene expression in different tissue samples,UBC30,Tub a and CyP were considered as the tissue samples was compared,followed by Cyp.For the biotic most suitable set ofreference genes.CyP and Tub a were the most stress treated samples,CyP and Tub_a showed lowest M values. stable reference genes in the biotic stress samples.For the abiotic The UBC30 and EF/-o genes displayed highest stability under stress treated samples,the gene set UBC30.EFI-a and ACTIN abiotic stress conditions,followed by ACTIN. was selected for most optimal normalisation.Generally,these The geNorm software was also used to calculate the pair-wise results show that different experiments (here comparing variation(Vn/Vn+1)for the determination of the optimal number developmental stages,tissues or stress treatments)require sets of control genes between the sequential normalisation factors of different reference genes for normalisation. (NF)(NFn and NFn+1)(Fig.1).Data obtained from all 20 samples were analysed together and showed that the V value with the inclusion of a fifth gene (V5/6)was 0.146,which was lower NormFinder analysis than the arbitrary cut off value of 0.15.This means that the most The NormFinder program was used as a different means for suitable set of reference genes should contain at least five genes further validation of the data.There were clear differences in 0.45 V2/3 ▣V3/4 ☐V4/5 ▣V56 0.40 ■V6/7 ☐V7/8 ☐V8/9 mV9/10 题V10W11 目V11/12 ▣V1213 0.35 0.30 0.25 0.20 0.15 0.10 0.05 Total Develoomental Different tissues Biotic stress Abiotic stress stages Fig.1.Pairwise variation(V)measureofthe candidate reference genes.When Vn/Vn +1 <0.15,then the optima number of reference genes is N.The optimal positions are indicated with an asterisk for the four strategic groups and all 20 samples together

GAPDH, EF1-a and UBC30. Furthermore, the genes ACTIN, CyP and GAPDH with lowest M values displayed less variation (from 0.1 to 0.3) than M values of other genes. UBC30 and Tub_a were the most stable genes when gene expression in different tissue samples was compared, followed by CyP. For the biotic stress treated samples, CyP and Tub_a showed lowest M values. The UBC30 and EF1-a genes displayed highest stability under abiotic stress conditions, followed by ACTIN. The geNorm software was also used to calculate the pair-wise variation (Vn/Vn + 1) for the determination of the optimal number of control genes between the sequential normalisation factors (NF) (NFn and NFn + 1) (Fig. 1). Data obtained from all 20 samples were analysed together and showed that the V value with the inclusion of a fifth gene (V5/6) was 0.146, which was lower than the arbitrary cut off value of 0.15. This means that the most suitable set of reference genes should contain at least five genes for accurate normalisation, i.e. ACTIN, CyP, UBC30, EF1-a and UBQ. When comparing developmental stages, ACTIN and CyP formed the optimal set of reference genes, whereas in different tissue samples, UBC30, Tub_a and CyP were considered as the most suitable set of reference genes.CyPandTub_awerethe most stable reference genes in the biotic stress samples. For the abiotic stress treated samples, the gene set UBC30, EF1-a and ACTIN was selected for most optimal normalisation. Generally, these results show that different experiments (here comparing developmental stages, tissues or stress treatments) require sets of different reference genes for normalisation. NormFinder analysis The NormFinder program was used as a different means for further validation of the data. There were clear differences in Table 3. Expression stability values of reference genes ranked by geNorm and NormFinder for the four strategic groups and all 20 samples together (ranking in parentheses) Gene geNorm NormFinder Total Developmental stages Different tissues Biotic stress Abiotic stress Total Developmental stages Different tissues Biotic stress Abiotic stress ACTIN 0.61 (1) 0.12 (1) 0.56 (4) 0.01 (1) 0.46 (2) 0.09 (1) 0.20 (3) 0.17 (1) 0.05 (5) 0.26 (3) ACTIN-1 1.21 (8) 0.36 (5) 0.52 (3) 0.71 (12) 0.51 (3) 0.90 (9) 0.34 (7) 0.46 (3) 0.74 (13) 0.32 (5) ACTIN-2 1.11 (7) 0.53 (7) 0.82 (7) 0.43 (8) 0.65 (7) 0.85 (8) 0.33 (6) 0.56 (6) 0.40 (7) 0.57 (10) GAPDH 1.02 (6) 0.20 (2) 1.03 (9) 0.31 (7) 0.70 (8) 0.68 (5) 0.29 (5) 0.69 (9) 0.56 (10) 0.50 (8) CyP 0.61 (1) 0.12 (1) 0.45 (2) 0.12 (5) 0.54 (4) 0.50 (4) 0.05 (1) 0.41 (2) 0.02 (3) 0.32 (6) EF1-a 0.74 (3) 0.23 (3) 0.66 (5) 0.22 (6) 0.28 (1) 0.39 (2) 0.28 (4) 0.51 (4) 0.40 (8) 0.32 (4) Tub_a 0.92 (5) 0.77 (9) 0.31 (1) 0.01 (1) 0.60 (6) 0.85 (7) 0.87 (11) 0.53 (5) 0.04 (4) 0.34 (7) 18S rRNA 1.43 (10) 1.54 (12) 1.22 (11) 0.58 (10) 0.57 (5) 1.17 (12) 3.44 (13) 1.03 (12) 0.62 (11) 0.18 (1) UBC30 0.72 (2) 0.34 (4) 0.31 (1) 0.061 (2) 0.28 (1) 0.46 (3) 0.52 (8) 0.62 (8) 0.00 (1) 0.21 (2) UBQ 0.86 (4) 0.66 (8) 0.91 (8) 0.07 (3) 0.74 (9) 0.79 (6) 0.59 (9) 0.81 (10) 0.00 (2) 0.54 (9) PPR 1.31 (9) 0.84 (10) 0.77 (6) 0.08 (4) 0.95 (12) 1.05 (10) 0.97 (12) 0.60 (7) 0.11 (6) 0.91 (13) PP2A 1.90 (12) 0.44 (6) 1.54 (12) 0.50 (9) 0.80 (10) 2.51 (13) 0.05 (2) 2.21 (13) 0.47 (9) 0.68 (11) MDH 1.56 (11) 0.92 (11) 1.12 (10) 0.65 (11) 0.86 (11) 1.16 (11) 0.68 (10) 0.83 (11) 0.73 (12) 0.68 (12) 0.45 Pairwise variation (V) V2/3 V6/7 V10/11 V3/4 V7/8 V11/12 V4/5 V8/9 V12/13 V5/6 0.40 V9/10 0.35 0.30 0.25 * * * * * 0.20 0.15 0.10 0.05 0 Total Developmental stages Different tissues Biotic stress Abiotic stress Fig. 1. Pairwise variation (V) measure of the candidate reference genes.When Vn/Vn + 1 <0.15,then the optimal number of reference genes is N. The optimal positions are indicated with an asterisk for the four strategic groups and all 20 samples together. Validation of reference genes in a wide of samples Functional Plant Biology 347

348 Functional Plant Biology D.Xiao et al. ranking the putative reference genes based on expression stability treatments when calculated using geNorm software,whereas in when comparing the two programs.However,both programs NormFinder analysis,also CyP and UBC30 were identified as identified the same genes as the most and the least stable reference the two most stable genes.Programs gave different output for genes (Table 3).For example,ACTIN was ranked as best ACTIN.which ranked fourth when comparing expression in reference gene when all tested samples were considered,by different tissues according to the geNorm software,although both programs geNorm and NormFinder,whereas Cyp was it was the best reference gene according to the NormFinder evaluated as the most stable reference gene when comparing software;however in abiotic stress,this gene ranked second different developmental stages only.An exception was /8SrRNA and third when analysed with geNorm and NormFinder that ranked at the top in abiotic stress treatment comparisons when respectively.The essential difference between geNorm and data were analysed using NormFinder,whereas it ranked sixth NormFinder is that with the software geNorm expression using geNorm. stability for each gene is determined by pair-wise comparison Furthermore,to present expression levels of each of the with all other reference genes across all experimental conditions, reference genes,the average of expression levels of /8S rRNA whereas NormFinder calculates the expression stability of a was used as reference to calculate the relative expression level gene per see,as a direct measure for its estimated expression (Fig.2).The results showed that ACTIN-/was the gene with the variation,without considering other genes tested. lowest average expression level,whereas Cyp had the highest 18S rRNA is a commonly accepted reference gene.However, average expression level. many reports have demonstrated that the expression of this gene varies under different experimental conditions (Vandesompele Discussion etal.2002;Jain et al.2006;Paolacciet al.2009).This may partly be explained by the fact that housekeeping genes are not only Real-time quantitative PCR has become a widespread approach to implicated in the basal cell metabolism but also participate in analyse gene expression in plant species.However,no single gene other cellular functions(Singh and Green 1993).In this study, has a constant expression level under all the tested experimental for all strategic groups except for the abiotic stress group,18S situations.Consequently,normalising the gene expression with rRNA ranked as the least stable gene.However,when comparing one reference gene under different experimental conditions different abiotic stress samples,it ranked 5th using geNorm will lead to biased results.In order to obtain more reliable and first using NormFinder.We have no explanation for this results from qRT-PCR experiments,it is crucial to select one observation,but will not suggest this gene. or a set of suitable reference genes. In grape,the genes GAPDH,ACTIN,EFI_a and SAND were The present study analysed the gene expression of 13 proposed as most relevant reference genes for normalisation candidate reference genes in a set of non-heading Chinese of qRT-PCR during berry development (Reid et al.2006). cabbage samples using geNorm and NormFinder.This A combination of UBO5 and EFI a was found as the most comparison showed some discrepancies in the ranking of the stable set of reference genes when comparing different candidate reference genes and in the identification of the best developmental stages in rice (Jain et al.2006).In poplar,a ones calculated by the two programs.However,there was combination of UBO,Tub_a and UBC was suggested as substantial agreement if the grouping of the genes with the reference gene set when comparing gene expression in 10 most and least stable expression was considered (Table 3). different tissues (Brunner et al.2004).We found that ACTIN For example,UBC30,Tub a and CyP were the most stably and CyP were the optimal combination when comparing gene expressed genes in different tissues and after biotic stress expression profiles from different developmental stages in non- 10.00 1.00 0.10 0.01 0.00 手S0叶 18S ACTIN-1 Fig.2.The expression level of the reference genes relative to 1&SrRNA in20 different samples.The mean Ct values of the 20 different samples were used for the relative expression analysis

ranking the putative reference genes based on expression stability when comparing the two programs. However, both programs identified the same genes as the most and the least stable reference genes (Table 3). For example, ACTIN was ranked as best reference gene when all tested samples were considered, by both programs geNorm and NormFinder, whereas Cyp was evaluated as the most stable reference gene when comparing different developmental stages only. An exception was 18S rRNA that ranked at the top in abiotic stress treatment comparisons when data were analysed using NormFinder, whereas it ranked sixth using geNorm. Furthermore, to present expression levels of each of the reference genes, the average of expression levels of 18S rRNA was used as reference to calculate the relative expression level (Fig. 2). The results showed that ACTIN-1 was the gene with the lowest average expression level, whereas Cyp had the highest average expression level. Discussion Real-time quantitative PCR has become a widespread approach to analyse gene expression in plant species. However, no single gene has a constant expression level under all the tested experimental situations. Consequently, normalising the gene expression with one reference gene under different experimental conditions will lead to biased results. In order to obtain more reliable results from qRT–PCR experiments, it is crucial to select one or a set of suitable reference genes. The present study analysed the gene expression of 13 candidate reference genes in a set of non-heading Chinese cabbage samples using geNorm and NormFinder. This comparison showed some discrepancies in the ranking of the candidate reference genes and in the identification of the best ones calculated by the two programs. However, there was substantial agreement if the grouping of the genes with the most and least stable expression was considered (Table 3). For example, UBC30, Tub_a and CyP were the most stably expressed genes in different tissues and after biotic stress treatments when calculated using geNorm software,whereas in NormFinder analysis, also CyP and UBC30 were identified as the two most stable genes. Programs gave different output for ACTIN, which ranked fourth when comparing expression in different tissues according to the geNorm software, although it was the best reference gene according to the NormFinder software; however in abiotic stress, this gene ranked second and third when analysed with geNorm and NormFinder respectively. The essential difference between geNorm and NormFinder is that with the software geNorm expression stability for each gene is determined by pair-wise comparison with all other reference genes across all experimental conditions, whereas NormFinder calculates the expression stability of a gene per see, as a direct measure for its estimated expression variation, without considering other genes tested. 18S rRNA is a commonly accepted reference gene. However, many reports have demonstrated that the expression of this gene varies under different experimental conditions (Vandesompele et al. 2002; Jain et al. 2006; Paolacci et al. 2009). This may partly be explained by the fact that housekeeping genes are not only implicated in the basal cell metabolism but also participate in other cellular functions (Singh and Green 1993). In this study, for all strategic groups except for the abiotic stress group, 18S rRNA ranked as the least stable gene. However, when comparing different abiotic stress samples, it ranked 5th using geNorm and first using NormFinder. We have no explanation for this observation, but will not suggest this gene. In grape, the genes GAPDH, ACTIN, EF1_a and SAND were proposed as most relevant reference genes for normalisation of qRT–PCR during berry development (Reid et al. 2006). A combination of UBQ5 and EF1_a was found as the most stable set of reference genes when comparing different developmental stages in rice (Jain et al. 2006). In poplar, a combination of UBQ, Tub_a and UBC was suggested as reference gene set when comparing gene expression in 10 different tissues (Brunner et al. 2004). We found that ACTIN and CyP were the optimal combination when comparing gene expression profiles from different developmental stages in non- 10.00 Relative expression 1.00 0.10 0.01 0.00 Cyp ACTIN MDH GAPDH EF 1-α 18S rRNA UBQ UBC30 Tub_α ACTIN-2 PP2A PPR ACTIN-1 Fig. 2. The expression level of the reference genes relative to 18S rRNA in 20 different samples. The mean Ct values of the 20 different samples were used for the relative expression analysis. 348 Functional Plant Biology D. Xiao et al

Validation of reference genes in a wide of samples Functional Plant Biology 349 heading Chinese cabbage,whereas ACTIN,CyP.UBC30,EFI-a the non-heading Chinese cabbage,analysed in this paper.In and UBO were the most stable reference genes when comparing another project conducted in our laboratory,the expression gene expression profiles of all 20 samples together. of all the 13 reference genes was tested in seven different Previously,most studies have recommended ACTIN tissue types,including whole plant of seedling,seed,stem,leaf expression as a reliable normalisation factor (Jian et al.2008) (young and old),turnip/root and flowers from two turnip However,in another study in B.rapa where expression in tissue genotypes(gene bank accessions CGN06678 and CGN07223). samples ofChinese cabbage were compared,ACTINwas not most harvested at five different developmental stages.The results stably expressed(Qi et al.2010).Jian et al.(2008)indicated that indicated that ACTIN and EF/-o.were the most stable different paralogues from the same gene family can have varying expressed genes in these tissues/developmental stages (data expression levels in different developmental stages of soybean not shown).We also conducted a gene expression profiling Our results showed that ACTIN was more stably expressed than experiment using a Nimble gene 300K array (http://www. ACTIN-/and ACTIN-2 in all four strategic groups tested and in ggbio.com/ggb/sub_contents.php?menu_id=2&sub_menu_id=0) all 20 samples analysed together.A previous observation in using 3-week-old Chinese cabbage seedlings grown under short Arabidopsis and Brachypodium distachyon showed that UBC (8 h light)or long(16 h light)daylengths with RNA extracted from had a very stable expression pattern (Czechowski et al.2005; leaves at 9 and at 21 h after dawn (see Table S1,available as Hong et al.2008).Our study also showed that the UBC30 Supplementary Material to this paper).The stability of the 13 exhibited very good expression stability when comparing its reference genes used in this study corresponded well with the expression in 20 samples and in biotic and abiotic stress results of the microarray experiment comparing growth under treated samples.The data presented in Czechowski et al. different daylengths with leaves sampled at different times of (2005),were screened for the 13 tested candidate reference the day.MDH and GAPDH were least stably expressed with genes of this study.This revealed that the transcript levels of expression levels differing at least 2-fold,whereas ACTIN and the five genes PP2A,UBO,PPR,Tub_a.and ACTIN,were very Cyp were very stable expressed similar to the results of this study. stable in developmental series and additional EF/-o was The results presented in this paper provide valuable identified as the best reference gene in abiotic stress information for future selection of reference genes in gene comparisons (Czechowski et al.2005).The EFI-a gene was expression studies in B.rapa crops,particularly non-heading also most stably expressed together with UBC30 and ACTIN in Chinese cabbage.We have identified distinct sets of genes our abiotic stress samples(Table 3).When we analysed only the appropriate for qRT-PCR in studies on plant developmental subgroup of hormone treated samples from the abiotic treatment stages,different tissues,biotic and abiotic stresses.We group,GAPDH was identified as the best ranking reference gene recommend at least five reference genes (ACTIN,Cyp, (results not shown)by both programs.Furthermore,when cold UBC30,EFI-a and UBO)for accurate gRT-PCR data and heat shock samples were analysed,GAPDH ranked first by normalisation,when studying gene expression across diverse geNorm (data not shown).However,when salt and drought types of samples.The genes ACTIN and CyP are the best choices stressed samples were analysed,GAPDH ranked last by both when studying expression in different developmental stages, programs(data not shown).This last finding is in sharp contrast whereas the genes Cyp,Tub_o.and UBC30 are preferred when with another study in heading Chinese cabbage,where GAPDH comparing gene expression in different tissues.When studying ranked as most stable reference gene during drought stress the effect ofbiotic stresses,CyP and Tub_aare recommended and (Qi et al.2010). UBC30,EF/-o and ACTINare proposed as reference genes when Recently,Hong et al.(2008)also showed that EF/-o was the studying abiotic stress.We conclude that different studies need most stably expressed gene when comparing Heat/Cold treated different sets of reference genes. Brachypodium distachyon,which was confirmed in our study by its stable expression in the abiotic stress group when analysed Acknowledgements by geNorm.Qi et al.(2010)reported that the gene CyP was not the best reference gene under drought and downy mildew stress This study was supported by the earmarked fund for the Priority Academic in Chinese cabbage.However,Cyp was identified as the most Program Development of Jiangsu Higher Education Institutions,China Agriculture Research System CARS-25-A-12 and the Fundamental stable reference gene under two temperature treatments Research Funds for the Central Universities KYZ200912.We thank (20/25C)in seagrass (Ransbotyn and Reusch 2006).In our Shuhang Wang and Dido Umer for sharing their data on expression of the study,Cyp is a stable reference gene in pakchoi for most studied gene set in turnip and yellow sarson and we thank Dr Yoonkang Hur strategic groups,except for the abiotic stress group. for performing the Nimblegen experiments. The present study indicates that none of the genes tested had a uniform expression profile in all 20 samples (tissues References developmental stages,stress responses),which stresses the fact that more reference genes need to be included when diverse sets Andersen CL,Jensen JL,Orntoft TF (2004)Normalization of real-time of samples are compared,as suggested by the geNorm software. quantitative reverse transcription-PCR data:a model-based variance Thus,before studying gene expression by qRT-PCR,it is estimation approach to identify genes suited for normalization,applied to bladder and colon cancer data sets.Cancer Research 64(15). necessary to consider a suitable set of reference genes. 5245-5250.do:10.1158/0008-5472.CAN-04-0496 We also tested these genes in different B.rapa accessions Argyropoulos D,Psallida C,Spyropoulos CG(2006)Generic normalization that represent different morpho types.Both in yellow sarson method for real-time PCR:application for the analysis of the mannanase (annual oil crop)and fodder tumip,the primers amplified PCR gene expressed in germinating tomato seed.FEBS Journal 273(4). products with similar efficiency(data not shown)compared with 770-777.doi:10.1111万.1742-4658.2006.05109.x

heading Chinese cabbage, whereas ACTIN, CyP, UBC30, EF1-a and UBQ were the most stable reference genes when comparing gene expression profiles of all 20 samples together. Previously, most studies have recommended ACTIN expression as a reliable normalisation factor (Jian et al. 2008). However, in another study in B. rapa where expression in tissue samples ofChinese cabbage were compared,ACTINwas not most stably expressed (Qi et al. 2010). Jian et al. (2008) indicated that different paralogues from the same gene family can have varying expression levels in different developmental stages of soybean. Our results showed that ACTIN was more stably expressed than ACTIN-1 and ACTIN-2 in all four strategic groups tested and in all 20 samples analysed together. A previous observation in Arabidopsis and Brachypodium distachyon showed that UBC had a very stable expression pattern (Czechowski et al. 2005; Hong et al. 2008). Our study also showed that the UBC30 exhibited very good expression stability when comparing its expression in 20 samples and in biotic and abiotic stress treated samples. The data presented in Czechowski et al. (2005), were screened for the 13 tested candidate reference genes of this study. This revealed that the transcript levels of the five genes PP2A, UBQ, PPR, Tub_a and ACTIN, were very stable in developmental series and additional EF1-a was identified as the best reference gene in abiotic stress comparisons (Czechowski et al. 2005). The EF1-a gene was also most stably expressed together with UBC30 and ACTIN in our abiotic stress samples (Table 3). When we analysed only the subgroup of hormone treated samples from the abiotic treatment group, GAPDH was identified as the best ranking reference gene (results not shown) by both programs. Furthermore, when cold and heat shock samples were analysed, GAPDH ranked first by geNorm (data not shown). However, when salt and drought stressed samples were analysed, GAPDH ranked last by both programs (data not shown). This last finding is in sharp contrast with another study in heading Chinese cabbage, where GAPDH ranked as most stable reference gene during drought stress (Qi et al. 2010). Recently, Hong et al. (2008) also showed that EF1-a was the most stably expressed gene when comparing Heat/Cold treated Brachypodium distachyon, which was confirmed in our study by its stable expression in the abiotic stress group when analysed by geNorm. Qi et al. (2010) reported that the gene CyP was not the best reference gene under drought and downy mildew stress in Chinese cabbage. However, CyP was identified as the most stable reference gene under two temperature treatments (20/25
C) in seagrass (Ransbotyn and Reusch 2006). In our study, CyP is a stable reference gene in pakchoi for most strategic groups, except for the abiotic stress group. The present study indicates that none of the genes tested had a uniform expression profile in all 20 samples (tissues, developmental stages, stress responses), which stresses the fact that more reference genes need to be included when diverse sets of samples are compared, as suggested by the geNorm software. Thus, before studying gene expression by qRT–PCR, it is necessary to consider a suitable set of reference genes. We also tested these genes in different B. rapa accessions that represent different morpho types. Both in yellow sarson (annual oil crop) and fodder turnip, the primers amplified PCR products with similar efficiency (data not shown) compared with the non-heading Chinese cabbage, analysed in this paper. In another project conducted in our laboratory, the expression of all the 13 reference genes was tested in seven different tissue types, including whole plant of seedling, seed, stem, leaf (young and old), turnip/root and flowers from two turnip genotypes (gene bank accessions CGN06678 and CGN07223), harvested at five different developmental stages. The results indicated that ACTIN and EF1-a were the most stable expressed genes in these tissues/developmental stages (data not shown). We also conducted a gene expression profiling experiment using a Nimble gene 300K array (http://www. ggbio.com/ggb/sub_contents.php?menu_id=2&sub_menu_id=0) using 3-week-old Chinese cabbage seedlings grown under short (8 h light) or long (16 h light) daylengths with RNA extracted from leaves at 9 and at 21 h after dawn (see Table S1, available as Supplementary Material to this paper). The stability of the 13 reference genes used in this study corresponded well with the results of the microarray experiment comparing growth under different daylengths with leaves sampled at different times of the day. MDH and GAPDH were least stably expressed with expression levels differing at least 2-fold, whereas ACTIN and Cyp were very stable expressed similar to the results of this study. The results presented in this paper provide valuable information for future selection of reference genes in gene expression studies in B. rapa crops, particularly non-heading Chinese cabbage. We have identified distinct sets of genes appropriate for qRT–PCR in studies on plant developmental stages, different tissues, biotic and abiotic stresses. We recommend at least five reference genes (ACTIN, CyP, UBC30, EF1-a and UBQ) for accurate qRT–PCR data normalisation, when studying gene expression across diverse types of samples. The genes ACTIN and CyP are the best choices when studying expression in different developmental stages, whereas the genes CyP, Tub_a and UBC30 are preferred when comparing gene expression in different tissues. When studying the effect of biotic stresses,CyPand Tub_aare recommended and UBC30,EF1-aandACTINare proposed as reference genes when studying abiotic stress. We conclude that different studies need different sets of reference genes. Acknowledgements This study was supported by the earmarked fund for the Priority Academic Program Development of Jiangsu Higher Education Institutions, China Agriculture Research System CARS-25-A-12 and the Fundamental Research Funds for the Central Universities KYZ200912. We thank Shuhang Wang and Dido Umer for sharing their data on expression of the studied gene set in turnip and yellow sarson and we thank Dr Yoonkang Hur for performing the Nimblegen experiments. References Andersen CL, Jensen JL, Ørntoft 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 Research 64(15), 5245–5250. doi:10.1158/0008-5472.CAN-04-0496 Argyropoulos D, Psallida C, Spyropoulos CG (2006) Generic normalization method for real-time PCR: application for the analysis of the mannanase gene expressed in germinating tomato seed. FEBS Journal 273(4), 770–777. doi:10.1111/j.1742-4658.2006.05109.x Validation of reference genes in a wide of samples Functional Plant Biology 349

350 Functional Plant Biology D.Xiao et al. Basa B.Solti A.Sarvari E,Tamas L (2009)Housekeeping gene selection Ramakers C,Ruijter JM,Deprez RHL,Moorman AFM(2003)Assumption- in poplar plants under Cd-stress:comparative study for real-time PCR free analysis of quantitative real-time polymerase chain reaction (PCR) normalisation.Functional Plant Biology 36(12).1079-1087. data.Neuroscience Letters 339(1).62-66.doi:10.1016/S0304-3940(02) doi:10.1071/FP09073 01423-4 Brunner A,Yakovlev I,Strauss S (2004)Validating intemal controls for Ransbotyn V,Reusch TBH (2006)Housekeeping gene selection for quantitative plant gene expression studies.BMC Plant Biology 4(1).14. quantitative real-time PCR assays in the seagrass Zosters marina doi:10.1186/1471-2229-4-14 subjected to heat stress.Limnology and Oceanography.Methods Chen X,Truksa M,Shah S,Weselake RJ(2010)A survey of quantitative real- 4N0V,367-373.doi:10.4319/1om2006.4.367 time polymerase chain reaction intemal reference genes for expression ReidK,Olsson N,Schlosser J,Peng F,Lund S(2006)An optimized grapevine studies in Brassica napus.Analytical Biochemistry 405(1),138-140. RNA isolation procedure and statistical determination of reference genes doi:10.1016f.ab.2010.05.032 for real-time RT-PCR during berry development.BMC Plant Biology Czechowski T.Stitt M.Altmann T.Udvardi MK.Scheible WR (2005) 6(1),27.doi:10.1186W/1471-2229-6-27 Genome-wide identification and testing of superior reference genes for Remans T,Smeets K,Opdenakker K,Mathijsen D,Vangronsveld J, transcript normalization in Arabidopsis.Plant Physiology 139(1),5-17. Cuypers A(2008)Normalisation of real-time RT-PCR gene expression doi:10.1104/pp.105.063743 measurements in Arabidopsis thaliana exposed to increased metal Ding J,Jia J,Yang L,Wen H,Zhang C,Liu W,Zhang D(2004)Validation of concentrations.Planta227(6,1343-1349.doi:10.1007/s00425-008- a rice specific gene,sucrose phosphate synthase,used as the endogenous 0706-4 reference gene for qualitative and real-time quantitative PCR detection Silver N,Best S,Jiang J,Thein S(2006)Selection of housekeeping genes of transgenes.Journal of Agricultural and Food Chemistry 52(11), for gene expression studies in human reticulocytes using real-time PCR 3372-3377.doi:10.1021/0f049915d BMC Molecular Biology 7(1),33.doi:10.1186/1471-2199-7-33 Hong S-Y,Seo P,Yang M-S,Xiang F,Park C-M(2008)Exploring valid Singh R,Green M (1993)Sequence-specific binding of transfer RNA reference genes for gene expression studies in Brachypodium distachyon by glyceraldehyde-3-phosphate dehydrogenase.Science 259(5093). by real-time PCR.BMC Plant Biology8(1),112.doi:10.1186/1471-2229- 365-368.doi:10.1126/science..8420004 8-112 Terrier N,Glissant D,Grimplet J,Barrieu F,Abbal P,Couture C,Ageorges A, Jain M,Nijhawan A,Tyagi A,Khurana J (2006)Validation of Atanassova R,Leon C,Renaudin J-P,Dedaldechamp F,Romieu C.Delrot housekeeping genes as internal control for studying gene expression S,Hamdi S (2005)Isogene specific oligo arrays reveal multifaceted in rice by quantitative real-time PCR.Biochemical and Biophysical changes in gene expression during grape berry (Vitis vinifera L.) Research Communications 345,646-651.doi:10.1016/j.bbrc.2006. development.P1amta222(5),832-847.doi10.1007/s00425-005-0017-y 04.140 The Brassica rapa Genome Sequencing Project Consortium (2011)The Jian B,Liu B,BiY.Hou W.Wu C,Han T(2008)Validation ofinternal control genome of the mesopolyploid crop species Brassica rapa.Nature for gene expression study in soybean by quantitative real-time PCR.BMC Genetics43,1035-1039.doi:10.1038/hg.919 Molecular Biology 9,59. Vandesompele J,De Preter K,Pattyn F,Poppe B,Van Roy N,De Paepe A, Kim BR,Nam HY,Kim SU,Kim SI,Chang YJ (2003)Normalization of Speleman F (2002)Accurate normalization of real-time quantitative reverse transcription quantitative-PCR with housekeeping genes in rice. RT-PCR data by geometric averaging of multiple internal control Biotechnology Letters25(21),1869-1872.doi:10.1023/A:10262980 genes.Genome Biology 3(7),research0034-research0034.doi:10.1186/ 32009 gb-2002-.3.-7-research0034 Lekanne Deprez RH,Fijnvandraat AC,Ruijter JM,Moorman AFM(2002) Wierschke S.Gigout S.Hor P,Lehmann TN,Dehnicke C,Brauer AU,Deisz Sensitivity and accuracy of quantitative real-time polymerase chain RA (2010)Evaluating reference genes to normalize gene expression reaction using SYBR green I depends on cDNA synthesis conditions. in human epileptogenic brain tissues.Biochemical and Biophysical Analytical Biochemistry 307(1),63-69.doi:10.1016/S0003-2697(02) Research Communications 403(3-4).385-390.doi:10.1016/.bbrc.2010. 00021-0 10.138 Lovdal T,Lillo C(2009)Reference gene selection for quantitative real-time Zhao J,Paulo M-J,Jamar D,Lou P,van Eeuwijk F,Bonnema G,Vreugdenhil PCR normalization in tomato subjected to nitrogen,cold,and light stress. D.Koornneef M(2007)Association mapping of leaf traits,flowering Analytical Biochemistry 387(2),238-242.doi:10.1016/j.ab.2009.01.024 time,and phytate content in Brassica rapa.Genome 50(10).963-973. Martinez-Beamonte R,Navarro MA,Larraga A,Strunk M,Barranquero C, doi:10.1139/G07-078 Acin S,Guzman MA,Inigo P,Osada J(2011)Selection ofreference genes Zhao J,Artemyeva A,Del Carpio DP,Basnet RK,Zhang N,Gao J,Li F, for gene expression studies in rats.Journal of Biotechnology 151(4), Bucher J,Wang X,Visser RG,Bonnema G(2010)Design of a Brassica 325-334.doi:10.1016 j.jbiotec.2010.12.017 rapa core collection for association mapping studies.Genome 53(11), Paolacci A.Tanzarella O.Porceddu E.Ciaffi M(2009)Identification and 884-898. validation of reference genes for quantitative RT-PCR normalization in wheat.BMC Molecular Biology 10(1),11.doi:10.1186/1471-2199-10-11 Qi J,Yu S,Zhang F,Shen X,Zhao X.Yu Y.Zhang D(2010)Reference gene selection for real-time quantitative polymerase chain reaction of mRNA transcript levels in Chinese cabbage (Brassica rapa L.ssp.pekinensis). Plant Molecular Biology Reporter 28(4),597-604.doi:10.1007/s11105- 010-0185-1 www.publish.csiro.au/journals/fpb

Basa B, Solti Á, Sárvári É, Tamás L (2009) Housekeeping gene selection in poplar plants under Cd-stress: comparative study for real-time PCR normalisation. Functional Plant Biology 36(12), 1079–1087. doi:10.1071/FP09073 Brunner A, Yakovlev I, Strauss S (2004) Validating internal controls for quantitative plant gene expression studies. BMC Plant Biology 4(1), 14. doi:10.1186/1471-2229-4-14 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. Analytical Biochemistry 405(1), 138–140. doi:10.1016/j.ab.2010.05.032 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. Plant Physiology 139(1), 5–17. doi:10.1104/pp.105.063743 Ding J, Jia J, Yang L, Wen H, Zhang C, Liu W, Zhang D (2004) Validation of a rice specific gene, sucrose phosphate synthase, used as the endogenous reference gene for qualitative and real-time quantitative PCR detection of transgenes. Journal of Agricultural and Food Chemistry 52(11), 3372–3377. doi:10.1021/jf049915d Hong S-Y, Seo P, Yang M-S, Xiang F, Park C-M (2008) Exploring valid reference genes for gene expression studies in Brachypodium distachyon by real-time PCR.BMC Plant Biology 8(1), 112. doi:10.1186/1471-2229- 8-112 Jain M, Nijhawan A, Tyagi A, Khurana J (2006) Validation of housekeeping genes as internal control for studying gene expression in rice by quantitative real-time PCR. Biochemical and Biophysical Research Communications 345, 646–651. doi:10.1016/j.bbrc.2006. 04.140 Jian B, Liu B, Bi Y, HouW,Wu C, Han T (2008) Validation of internal control for gene expression study in soybean by quantitative real-time PCR. BMC Molecular Biology 9, 59. Kim BR, Nam HY, Kim SU, Kim SI, Chang YJ (2003) Normalization of reverse transcription quantitative-PCR with housekeeping genes in rice. Biotechnology Letters 25(21), 1869–1872. doi:10.1023/A:10262980 32009 Lekanne Deprez RH, Fijnvandraat AC, Ruijter JM, Moorman AFM (2002) Sensitivity and accuracy of quantitative real-time polymerase chain reaction using SYBR green I depends on cDNA synthesis conditions. Analytical Biochemistry 307(1), 63–69. doi:10.1016/S0003-2697(02) 00021-0 Løvdal T, Lillo C (2009) Reference gene selection for quantitative real-time PCR normalization in tomato subjected to nitrogen, cold, and light stress. Analytical Biochemistry 387(2), 238–242. doi:10.1016/j.ab.2009.01.024 Martínez-Beamonte R, Navarro MA, Larraga A, Strunk M, Barranquero C, Acín S, GuzmanMA, Iñigo P, Osada J (2011) Selection of reference genes for gene expression studies in rats. Journal of Biotechnology 151(4), 325–334. doi:10.1016/j.jbiotec.2010.12.017 Paolacci A, Tanzarella O, Porceddu E, Ciaffi M (2009) Identification and validation of reference genes for quantitative RT-PCR normalization in wheat. BMC Molecular Biology 10(1), 11. doi:10.1186/1471-2199-10-11 Qi J, Yu S, Zhang F, Shen X, Zhao X, Yu Y, Zhang D (2010) Reference gene selection for real-time quantitative polymerase chain reaction of mRNA transcript levels in Chinese cabbage (Brassica rapa L. ssp. pekinensis). Plant Molecular Biology Reporter 28(4), 597–604. doi:10.1007/s11105- 010-0185-1 Ramakers C, Ruijter JM, Deprez RHL, Moorman AFM (2003) Assumption￾free analysis of quantitative real-time polymerase chain reaction (PCR) data. Neuroscience Letters 339(1), 62–66. doi:10.1016/S0304-3940(02) 01423-4 Ransbotyn V, Reusch TBH (2006) Housekeeping gene selection for quantitative real-time PCR assays in the seagrass Zosters marina subjected to heat stress. Limnology and Oceanography, Methods 4(NOV), 367–373. doi:10.4319/lom.2006.4.367 Reid K, Olsson N, Schlosser J, Peng F, Lund S (2006) An optimized grapevine RNA isolation procedure and statistical determination of reference genes for real-time RT-PCR during berry development. BMC Plant Biology 6(1), 27. doi:10.1186/1471-2229-6-27 Remans T, Smeets K, Opdenakker K, Mathijsen D, Vangronsveld J, Cuypers A (2008) Normalisation of real-time RT-PCR gene expression measurements in Arabidopsis thaliana exposed to increased metal concentrations. Planta 227(6), 1343–1349. doi:10.1007/s00425-008- 0706-4 Silver N, Best S, Jiang J, Thein S (2006) Selection of housekeeping genes for gene expression studies in human reticulocytes using real-time PCR. BMC Molecular Biology 7(1), 33. doi:10.1186/1471-2199-7-33 Singh R, Green M (1993) Sequence-specific binding of transfer RNA by glyceraldehyde-3-phosphate dehydrogenase. Science 259(5093), 365–368. doi:10.1126/science.8420004 Terrier N, Glissant D, Grimplet J, Barrieu F, Abbal P, Couture C, Ageorges A, Atanassova R, Léon C, Renaudin J-P, Dédaldéchamp F, Romieu C, Delrot S, Hamdi S (2005) Isogene specific oligo arrays reveal multifaceted changes in gene expression during grape berry (Vitis vinifera L.) development. Planta 222(5), 832–847. doi:10.1007/s00425-005-0017-y The Brassica rapa Genome Sequencing Project Consortium (2011) The genome of the mesopolyploid crop species Brassica rapa. Nature Genetics 43, 1035–1039. doi:10.1038/ng.919 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 Biology 3(7), research0034–research0034. doi:10.1186/ gb-2002-3-7-research0034 Wierschke S, Gigout S, Horn P, Lehmann TN, Dehnicke C, Bräuer AU, Deisz RA (2010) Evaluating reference genes to normalize gene expression in human epileptogenic brain tissues. Biochemical and Biophysical Research Communications 403(3–4), 385–390. doi:10.1016/j.bbrc.2010. 10.138 Zhao J, Paulo M-J, Jamar D, Lou P, van Eeuwijk F, Bonnema G, Vreugdenhil D, Koornneef M (2007) Association mapping of leaf traits, flowering time, and phytate content in Brassica rapa. Genome 50(10), 963–973. doi:10.1139/G07-078 Zhao J, Artemyeva A, Del Carpio DP, Basnet RK, Zhang N, Gao J, Li F, Bucher J, Wang X, Visser RG, Bonnema G (2010) Design of a Brassica rapa core collection for association mapping studies. Genome 53(11), 884–898. 350 Functional Plant Biology D. Xiao et al. www.publish.csiro.au/journals/fpb