Identification and validation of reference genes for quantitative real-time PCR studies in Hedera helix L.

Plant Physiology and Biochemistry 108 (2016) 286e294

Contents lists available at ScienceDirect

Plant Physiology and Biochemistry journal homepage: www.elsevier.com/locate/plaphy

Research article

Identification and validation of reference genes for quantitative real-time PCR studies in Hedera helix L. Hua-peng Sun a, 1, Fang Li a, b, 1, Qin-mei Ruan a, Xiao-hong Zhong a, * a b

Horticulture & Landscape College, Hunan Agricultural University, Changsha, Hunan 410128, China National Center for Citrus Improvement, Changsha, Hunan 410128, China

a r t i c l e i n f o

a b s t r a c t

Article history: Received 21 April 2016 Received in revised form 21 July 2016 Accepted 21 July 2016 Available online 22 July 2016

Reference gene evaluation and selection are necessary steps in gene expression analysis, especially in new plant varieties, through reverse transcription quantitative real-time PCR (RT-qPCR). Hedera helix L. is an important traditional medicinal plant recorded in European Pharmacopoeia. Research on gene expression in H. helix has not been widely explored, and no RT-qPCR studies have been reported. Thus, it is important and necessary to identify and validate suitable reference genes to for normalizing RT-qPCR results. In our study, 14 candidate protein-coding reference genes were selected. Their expression stability in five tissues (root, stem, leaf, petiole and shoot tip) and under seven abiotic stress conditions (cold, heat, drought, salinity, UV-C irradiation, abscisic acid and methyl jasmonate) were evaluated using geNorm and NormFinder. This study is the first to evaluate the stability of reference genes in H. helix. The results show that different reference genes should be chosen for normalization on the basis of various experimental conditions. F-box was more stable than the other selected genes under all analysis conditions except ABA treatment; 40S was the most stable reference gene under ABA treatment; in contrast, EXP and UBQ were the most unstable reference genes. The expressions of HhSE and Hhb-AS, which are two genes related to the biosynthetic pathway of triterpenoid saponins, were also examined for reference genes in different tissues and under various cold stress conditions. The validation results confirmed the applicability and accuracy of reference genes. Additionally, this study provides a basis for the accurate and widespread use of RT-qPCR in selecting genes from the genome of H. helix. © 2016 Elsevier Masson SAS. All rights reserved.

Keywords: Reference gene Hedera helix L. RT-qPCR Differential expression Abiotic stress

1. Introduction Hedera helix L., a perennial evergreen climbing plants in the family Araliaceae, originated from Europe and is now cultivated worldwide. H. helix is often used as an ornamental plant for indoor and vertical gardening because of its evergreen colour, various leaf shapes and strong climbing ability. However, its role as a traditional medicinal plant is often disregarded (Landgrebe et al., 1999; Lutsenko et al., 2010). The fresh leaves and stems of H. helix are used to treat cough, asthma, bronchitis and other respiratory diseases (Hofmann et al., 2003; Lutsenko et al., 2010; Cwientzek et al., 2011). Its medicinal value and effectiveness have been accepted by several European countries; thus, this species is recorded in the European Pharmacopoeia as a herbal medicine (European

* Corresponding author. E-mail address: [email protected] (X.-h. Zhong). 1 These two authors contributed equally. http://dx.doi.org/10.1016/j.plaphy.2016.07.022 0981-9428/© 2016 Elsevier Masson SAS. All rights reserved.

Pharmacopoeia 7.0, 2010). The compounds extracted from this species have also been developed into tablet and syrup formulations (Khdair et al., 2010; Stauss-Grabo et al., 2011). As a medicinal plant, H. helix mainly contains triterpenoid saponins, which include hederacoside C, a-hederin, hederacoside B and hederacoside D (Ilhami et al., 2004; Gepdiremen et al., 2005). Although the pharmacodynamics of H. helix have been extensively investigated (Fazio et al., 2009; Mendel et al., 2011, 2013; Holzinger and Chenot, 2011), the biosynthesis of triterpenoid saponins in H. helix has yet to be reported. Thus far, molecular biology methods have been used to examine functional genes in medicinal plants (Yuan et al., 2008). Biological approaches primarily intend to identify the biosynthetic pathway of active ingredients in medicinal plants and to determine the functional genes and their expression patterns in these pathways; by using biological approaches, researchers can elucidate the regulatory mechanisms of these genes and examine genome diversity (Colebatch et al., 2002). Reverse transcription quantitative real-time PCR (RT-qPCR) is widely utilized to analyse gene relative expression levels under

H.-p. Sun et al. / Plant Physiology and Biochemistry 108 (2016) 286e294

different experimental treatments. This approach provides several advantages, including reproducibility, high sensitivity, accuracy and rapidness; however, its results are affected by the RNA quality, the expression effect of target genes and other factors that contribute to non-uniform test results (Bustin, 2002; Gachon et al., 2004; Derveaux et al., 2010). To obtain the true differences in expressions of target genes by RT-qPCR, we should select stably expressed reference genes for standard correction (Chervoneva et al., 2010). However, the ideal stably expressed reference genes under certain conditions have yet to be reported. The stability of reference genes can be changed using different samples, organs, developmental stages and physiological conditions (Bustin, 2009; Kozera and Rapacz, 2013). In lettuce, TUB and EIF4A1 are the most stable protein-coding reference genes under abscisic acid (ABA) treatment; EIF2A and TIP41 are the most stable genes under drought stress (Borowski et al., 2014). In developmental studies of grapevine berry, EF1-a and SAND were identified as the most stable reference genes (Reid et al., 2006). In grapevine leaf stress treatment, EF1-a, CYP and UBC are the highest scoring genes (Borges et al., 2014). Thus, we must select suitable and stable reference genes based on different experimental conditions and avoid the random use of one or more reference genes to obtain reliable real-time PCR results. Several medicinal plants, such as Panax ginseng (Liu et al., 2014a), Panax notoginseng (Wu et al., 2015), Catharanthus roseus (Pollier et al., 2014) and Atropa belladonna (Li et al., 2014), have been investigated; nevertheless, the selection and estimation of reference genes in H. helix remains unperformed. In molecular studies on H. helix, a multifactorial analysis is an essential prerequisite to evaluate the stability of reference genes. In our study, 14 candidate reference genes (ACT, GAPDH, 18S, 40S, UBQ, TUA, TUB, EF-1a, TIP41, EXP, CYP, F-box, PGK and PP2A) were selected based on our previous work to identify the most suitable reference genes for normalization of RT-qPCR data obtained from different H. helix tissues exposed to various abiotic stresses, including cold, heat, drought, salinity, UV-B, methyl jasmonate (MeJA) and abscisic acid (ABA). The expression stability of these reference genes was analysed using geNorm and NormFinder. The expression of two target genes, namely, HhSE (Han et al., 2010; Luo et al., 2011; Liu et al., 2014b; Ye et al., 2014) and Hhb-AS (Kushiro et al., 1998; Haralampidis et al., 2002), which are related to the biosynthesis of triterpenoid saponins in H. helix, was also examined to verify the reliability of the selected reference genes. Our data provide a reliable set of reference genes suitable for RT-PCR analysis in H. helix under different experimental conditions. 2. Materials and methods 2.1. Plant materials and treatments H. helix plants were cut in the greenhouse at the National Center for Citrus Improvement at Hunan Agricultural University, Hunan Province, China, on April 11, 2014. The wood cuttings were collected from the Hunan Research Institute of Vine Plants. One-year-old H. helix plant cuttings were used for the greenhouse experiment with completely randomized design and three biological replications. Each biological replicate consisted of 5e8 plants. For different tissue samples, fresh roots, stems, leaves, petiole and shoot tips were harvested, immediately frozen in liquid nitrogen and stored at
80  C for RNA extraction. In all treatments under abiotic stresses, the plants were grown in growth chambers (Life apparatus, Ningbo, China) with a 16/8-h photoperiod, PAR of 300 mmol/m2/s and relative humidity of 60%. Leaf tissues were harvested after treatments and frozen in liquid nitrogen and immediately stored at
80  C for RNA extraction. For the cold stress treatment, plants were placed at 4  C for 0, 5, 15 and 30 days. For

287

the heat stress treatment, plants were placed at 40  C for 2 days. For the drought stress treatment, plants were not watered for 20 days; at the end of this treatment, the measured soil cultivation water content was 5%. For the salinity stress treatment, plants were irrigated with 100 mM NaCl for 2 days. For the UV irradiation treatment, plants were exposed to UV-C radiation (Philips TUV 30 W, 92 mW/cm2 at 253 nm) at a distance of 15 cm from the source for 15 min and were then incubated in the dark for 2 days (Borges et al., 2014). For ABA treatment, plants were sprayed with 100 mM ABA for 0, 12, 24 and 48 h. For MeJA treatment, plants were sprayed with 100 mM MeJA for 0, 12, 24 and 48 h. 2.2. RNA isolation and cDNA synthesis Total RNA was extracted from all prepared samples with an RNAprep Pure Kit (Polysaccharides & Polyphenolics-rich; Tiangen, Beijing, China) and treated with RNase-free DNase I according to the manufacturer's instructions. The RNA concentration and purity were determined with a Nano Photometer P-Class instrument (Implen, Munich, Germany). The RNA extract had a 260/280 ratio between 1.9 and 2.1 and a 260/230 ratio of approximately 2.0. The RNA integrity was also checked on 1% agarose gels. Total RNA (1.0 mg) was used for reverse transcription with a FastQuant RT Kit (Tiangen, Beijing, China) in a 20 mL reaction volume according to the manufacturer's instructions. 2.3. Primer design and RT-qPCR conditions The sequences of all the 14 candidate reference genes, including traditional and novel protein-coding reference genes, were obtained from our transcriptome database as constructed by the Illumina Hiseq™ 2500 platform (1Gene, Hangzhou, China). Specific primer pairs were designed with the Beacon Designer 8 software according to the following parameters: primer sequences of 18e24 nucleotides, amplicon length of 75e150 bp, melting temperature (Tm) of 55e60  C and GC content of 40%e60%. All primer pairs were synthesized using a commercial supplier (Sangon, Shanghai, China) and tested by regular PCR. The products were analysed by electrophoresis on 1.0% agarose gels before RT-qPCR. In addition, the amplification efficiency (E) and correlation coefficient (R2) were calculated by a standard curve with a 5-fold serial dilution of mixed cDNA (1 mg/mL) (Bustin, 2009). The primer sequences, GeneBank accession numbers, amplicon length, Tm, GC content, E and R2 of the 14 candidate reference genes are listed in Table 1. RT-qPCR was performed in 96-well plates in a Bio-Rad CFX96 real-time PCR system (Bio-Rad, CA, USA) with a SYBR Green-based PCR assay. The final volume for each reaction was 20 mL with the following components: 2 mL diluted cDNA template (1 mg/mL), 10 mL SYBR Green Mix (Bio-Rad, CA, USA), 2.5 mL forward primer (2.5 mM), 2.5 mL reverse primer (2.5 mM) and 3 mL ddH2O. The reaction was conducted under the following conditions: 95  C for 3 min, followed by 40 cycles of denaturation at 95  C for 10 s and annealing/ extension at 56  C for 30 s. The melting curve was obtained by heating the amplicon from 65  C to 95  C with increments of 0.5  C in 5 s. Each RT-qPCR analysis was performed with three technical replicates. 2.4. Reference gene expression stability determination and statistical analysis The two most widely used statistical software programs, geNorm and NormFinder, were used to calculate and evaluate the stability of the 14 candidate reference genes under different experimental conditions. First, raw RT-qPCR data were obtained from the CFX manager (Bio-Rad) and exported into Excel

288

H.-p. Sun et al. / Plant Physiology and Biochemistry 108 (2016) 286e294

Table 1 Candidate reference genes and target genes description and primer sequences. Gene

GeneBank accession number

Primer sequences (forward/reverse)

Amplicon Tm ( C)

Amplicon length (bp)

E (%)

R2

ACT GAPDH 18S 40S TUA TUB UBQ EF-1a EXP PP2A PGK TIP41 CYP F-box HhSE Hhb-AS

KU942510 KU942515 KU942508 KU942509 KU942519 KU942520 KU942521 KU942512 KU942513 KU942517 KU942516 KU942518 KU942511 KU942514 KU942524 KU942522

CAGGAAGAGGAACATACT/AGATGGTTGGAATAGAGC GGTGTCAATGAGAAGGAATAC/TCAACAATGCCGAATCTG GCTCGTTATAGGACTTGAC/TTACCAGCCCTTGACATAT TGGTATCTTCAGCACTATGG/TTCGCTTCCTCAGTAACAG TATGGAGGAAGGAGAGTT/CCGTCTTCATCATCATCA CTTGCTGTGAACCTGATT/AATGTATTGCTGCGATCC GAGTCCACCCTTCATTTG/GTGATTGTCTTGCCAGTA CCATTGATATTGCCTTGTG/ATAAAGTCACGATGTCCAG TCACAATCAACGGTTTCC/CTCACCCTCACAATATCC GGTGCCAGACAACTAATAG/CCACATTCATCTCGCATAT GATGTGGTTATTGCTGAC/GTAGTATCCAACGCTTCA AACCACTTGCCAGAGATG/CTTCCAGCCAACTAGAGC TCATCAAGAAGCATACCG/AAGAACTGCGATCCATTG GCCTTAATCGGAATGAGA/CTTGTCCTTACTTGGCTT GTGGAGGAATGACTGTTG/AGTCGTGGAGATTGTGTA GTGTATGCTTGCTTGTTG/ATCGTCATTCCATCTTCAG

79.0 80.0 82.0 81.5 80.5 81.0 79.5 80.0 80.5 79.5 80.5 79.5 83.0 79.0 79.0 78.5

90 116 126 104 102 90 80 82 79 108 137 108 76 98 78 105

103.2 109.0 93.6 91.6 106.0 105.7 108.6 94.5 105.1 104.2 94.1 108.0 95.5 97.8 103.5 97.4

0.992 0.992 0.999 0.997 0.998 0.995 0.999 0.998 0.994 0.992 0.997 0.993 0.998 0.993 0.994 0.996

datasheets (Microsoft Excel 2013). The Ct values were used to determine the expression levels of candidate reference genes (Bustin, 2009). Before Ct values were entered into geNorm and NormFinder, all Ct values must be transformed into to relative quantification data. Each Ct value was transformed in to a “DCt” value with the highest DCt value of 0, while all other values are less than 0. For each data point, the equation 2(
DCt) was applied. Therefore, all data are expressed relative to the expression of the least expressed gene, which was used to measure the stability of the reference genes and which can be used to rank the candidate reference genes (Vandesompele et al., 2002). For geNorm, the expression stability value (M) of each reference gene was calculated based on the average pairwise variation (V) between all genes tested (Vandesompele et al., 2002). For NormFinder, an ANOVAbased model of each reference gene was used to calculate the expression stability value by determining the inter-and intra-group variations; the gene with the lowest value has the most stable expression (Claus et al., 2004). Finally, the most stable reference genes from H. helix can be obtained under different experimental conditions. 2.5. Validation of reference gene analysis HhSE and Hhb-AS are key enzyme-coding genes related to the biosynthesis of triterpenoid saponins in H. helix (Kushiro et al., 1998; Haralampidis et al., 2002; Han et al., 2010; Luo et al., 2011; Ye et al., 2014). The genes were cloned from our transcriptome database and used as target genes to confirm the reliability of the potential reference genes identified for RT-qPCR. The relative expression levels of HhSE and Hhb-AS under different tissues and cold stress treatment were determined and normalized according to the most and least stable reference genes according to geNorm and NormFinder in the same RT-qPCR conditions mentioned above.

ranged from 0.992 for ACT, GAPDH or PP2A to 0.999 for 18S and UBQ (Table 1). 3.2. Expression profile of the reference genes The raw expression levels of the 14 candidate reference genes were detected by RT-qPCR and analysed across all samples; lower Ct values represented higher expression levels (Fig. 1). The 14 tested reference genes indicated a relatively wide range of mean Ct values, from 15.53 in 18Se26.61 in EXP. All tested genes were grouped into two arbitrary categories: 10 genes with mean Ct values below 24 cycles (18S, CYP, 40S, TUA, UBQ, PGK, EF-1a, F-box, GAPDH and ACT) displayed high expression levels, whereas four genes with mean Ct values above 24 cycles (EXP, PP2A, TIP41 and TUB) displayed low expression levels. In addition, individual reference genes indicated different coefficients of variation (lower values represent less variability) under varying conditions, as shown in Fig. 1. PP2A (3.67 cycles) had the least variation whereas EXP (8.56 cycles) had the most variation; the other differentially expressed genes were TUA (3.68 cycles), 40S (4.27 cycles), F-box (4.34 cycles), 18S (4.46 cycles), EF-1a (4.52 cycles), ACT (4.57 cycles), TIP41 (5.01 cycles), CYP (5.50 cycles), UBQ (5.52 cycles), PGK (6.96 cycles), TUB (7.31 cycles) and GAPDH (7.47 cycles). 3.3. Expression stability of all candidate reference genes Given the wide range of variations among the 14 candidate reference genes in terms of their expression levels based on the raw Ct values, statistical methods that evaluate their stability were

35 30

3.1. Verification of amplicons, primer specificity and PCR amplification efficiency

Ct value

25

3. Results

20 15 10 5

The specific amplification of all primer pairs used for candidate reference genes was verified by regular PCR and RT-qPCR. Agarose gel electrophoresis indicated that all primer pairs amplified a PCR product (Fig. S1b), and the melting curve analysis validated that each primer pair presented a single peak (Fig. S1a). The amplification efficiency of RT-qPCR across all 14 reference genes varied from 91.6% for 40S to 109.0% for GAPDH; the correlation coefficients

0

ACT GAPDH 18S

40S

TUA TUB UBQ EF-1α EXP PP2A PGK TIP41 CYP F-box

Fig. 1. The RT-qPCR Ct values of different candidate reference genes. Expression data displayed as Ct values for each reference gene in all Hedera helix L. samples. The line across the box depicts the median. The box indicates the 25th and 75th percentiles, and whisker caps represent the maximum and minimum values. The lower the boxes and whisker, the smaller the variations.

H.-p. Sun et al. / Plant Physiology and Biochemistry 108 (2016) 286e294

necessary. Thus, the geNorm and NormFinder algorithms were used for further analysis. For geNorm analysis, all candidate reference genes were evaluated by the M values below the threshold of 1.5 (Vandesompele et al., 2002). The M value was calculated based on the average pairwise variation between all genes tested under different conditions and ranked according to the stepwise exclusion of the least stable gene; a lower M value represents higher gene expression stability (Vandesompele et al., 2002). Not all of the most stable reference genes in the 14 candidate reference genes were identical across different treatments. As shown in Fig. 2, the results obtained through geNorm implied that ACT and F-box were the most stable genes for all samples of different tissues and abiotic stress treatments. Furthermore, F-box and TIP41 were the most stable genes for total tissue samples, whereas ACT and F-box proved to be the most stable genes for total abiotic stress treatment samples. All genes were individually analysed under different types of abiotic stress as well. The most stable genes under cold, heat, drought and salinity stresses were ACT and PGK, GAPDH and TUA, F-box and TIP41, 40S and F-box, respectively. Under UV-C radiation, the most stable genes were PP2A and ACT. The most stable genes under the ABA treatment were 40S and PP2A, whereas, under MeJA treatment, CYP and PP2A were the most stable genes. Relatively, the least stable genes were as follow: EXP for the total samples, total tissues and the cold or ABA treatment; UBQ for the total abiotic stress conditions and heat, MeJA and UV-C treatment; TUA for drought; and TUB for salinity treatment. Furthermore, geNorm can calculate the pairwise variation Vn/ nþ1 to determine the optimal number of control genes for normalization. The ideal pairwise variation (V) score under 0.15 was recommend (Vandesompele et al., 2002; Jacinta et al., 2014); however, the proposed value of 0.15 must not to be regarded as a strict cut-off because it is only intended to guide the determination of the reference genes' optimal number(Marum et al., 2012; Chen et al., 2015). As shown in Fig. 3, the results of this work indicate that the use of two reference genes sufficiently normalizes the results of RT-qPCR analysis among all H. helix samples under different experimental conditions because all of the V2/3 values were ideal and lower than 0.15. According to geNorm, the best combinations were ACT þ F-box for total samples and total abiotic stresses, TIP41þF-box for total tissues and drought stress, ACT þ PGK for cold stress, GAPDH þ TUA for heat stress, 40S þ F-box for salinity stress, ACT þ PP2A for UV-C radiation, 40S þ PP2A for ABA treatment and PP2A þ CYP for MeJA treatment. NormFinder analysis is unlike geNorm because this ANOVAbased analysis considers intra- and inter-group variations to evaluate expression stability and to provide a direct measure of variation (Claus et al., 2004). NormFinder revealed that the 14 candidate reference genes were ranked under different experimental conditions relative to their stability values (Table 2); lower values indicated higher stability. Therefore, the top three most stable reference genes across the samples were F-box (0.222), EF1a (0.254) and ACT (0.255). F-box (0.165, 0.162) and ACT (0.167, 0.194) were ranked as the top two most stable genes in all of the tissues and abiotic stress samples. Under different abiotic stresses, F-box was the most stable gene with lowest value for cold (0.142), salinity (0.070), UV-C (0.065) and MeJA (0.111). TUA was the most stable gene for heat (0.041) and ABA (0.159), whereas TIP41 was the most stable gene for drought (0.180) treatment. Relatively, the gene that exhibited the maximum value was EXP for total samples (0.875), total tissues (0.677), cold (0.795) and ABA (0.562) treatments, UBQ for total abiotic stress (0.713), heat (0.788) and UV-C (0.284) treatments, TUA for drought (0.428) and MeJA (0.351) treatments and TUB for the salinity (0.410) treatment. Moreover, NormFinder analysed the best combination of two genes in terms

289

of the stability value. For total samples, the best combination was F-box þ EF-1a (0.200), ACT þ F-box (0.131) for total tissues samples, and F-box þ ACT (0.140) for total abiotic stress. Under different abiotic stress, PGK þ F-box (0.135) was for cold stress, GAPDH þ TUA (0.057) was for heat stress, TIP41þF-box (0.086) was for drought stress and F-boxþ40S (0.039) was for salinity stress, PP2A þ F-box (0.053) was for UV-C irradiation treatment, 40S þ TUA (0.085) was for ABA treatment and F-box þ GAPDH (0.059) was for MeJA treatment. To present the analysis results obtained by geNorm and NormFinder, the five most stable genes and best combination of two genes, as well as the least stable genes, are comprehensively listed in Table 3. 3.4. Reference gene validation The main active ingredients in H. helix are triterpenoid saponins. HhSE and Hhb-AS are the key enzymes in the biosynthetic pathway of triterpenoid saponins. Their expression levels may be directly related with the content of triterpenoid saponins and are the basis for our future biosynthesis regulation studies. The relative expression levels of these two genes (HhSE and Hhb-AS) were used to examine and normalize the results observed in different tissues and under cold stress to demonstrate the feasibility of the selected reference genes (Fig. 4). According to the comprehensive analysis of geNorm, NormFinder and the raw Ct value, two sets of reference genes were selected. The most stable reference genes were F-box, ACT and F-box þ ACT for different tissues and PGK, ACT and PGK þ ACT for cold stress; the most unstable gene was EXP for both. Under different tissues, the expression of HhSE and Hhb-AS in the stem was assumed as ‘1’ and we used 2(
DDCt) to calculate their relative expression levels in other samples. As shown in Fig. 4aeb, when the most stable reference genes (F-box, ACT, Fbox þ ACT) were used for normalization, there was no signification difference in their relative expression of both HhSE and Hhb-AS. However, when the least stable reference gene (EXP) was used for normalization, there was significant difference among their relative expression levels in leaf samples (p < 0.05). Similar results were observed under cold treatment: the expression of HhSE and Hhb-AS in 30 d was supposed as ‘1’ and we used 2(
DDCt) to calculate their relative expression levels in other samples. As shown in Fig. 4ced, when the most stable reference genes (PGK, ACT, PGK þ ACT) were used for normalization, there was no significant difference in the relative expression levels of both HhSE and Hhb-AS. When the least stable reference gene (EXP) was used for normalization, there was significant difference among their relative expression levels in 0 d (p < 0.01) and 5 d (p < 0.05) samples. 4. Discussion and conclusion No reports on reference genes in H. helix have been available prior to this work. Thus, we selected 14 candidate reference genes for evaluation by consulting the reference gene evaluation studies in other plant species as well as in our transcriptome database. Among the 14 genes, traditional housekeeping genes, such as ACT, GAPDH and 18S, and novel protein coding genes, such as PGK and Fbox, were used (Kozera and Rapacz, 2013; Borowski et al., 2014). The results of the regular PCR gel electrophoresis and the RT-qPCR dissolution curve showed that all of the primer pairs of the 14 candidate reference genes showed good specificity. The results of standard curve equation and amplification efficiency similarly proved that all of the designed primers were suitable for the reference gene selection and provided a good foundation for the selection and evaluation of H. helix reference genes.

290

H.-p. Sun et al. / Plant Physiology and Biochemistry 108 (2016) 286e294 1.2

Average expression stability (M)

0.9

Total

1.1

Tissues 0.8

1 0.9

0.7

0.8

0.6

0.7

0.5

0.6 0.4

0.5 0.4

0.3

0.3

0.2

0.2 0.1

0.1 0 EXP

UBQ

TUB

18S

GAPDH

PGK

TIP41

CYP

EF-1α

TUA

PP2A

0 EXP

ACT F-box

TUB

UBQ

18S

PGK

CYP

GAPDH

TUA

40S

EF-1α

PP2A

ACT

TIP41 F-box

0.7

1.1

Abiotic stresses

1

Average expression stability (M)

40S

Drought 0.6

0.9 0.8

0.5

0.7 0.4

0.6 0.5

0.3

0.4 0.2

0.3 0.2

0.1

0.1 0 UBQ

18S

EXP

Average expression stability (M)

1

TUB

PGK

GAPDH

TIP41

<::::: Least stable genes

CYP

TUA

PP2A

Most stable genes ::::>

EF-1α

40S

ACT F-box

Heat

0.8

0.8

0.7

0.7

0.6

0.6

0.5

0.5

0.4

0.4

0.3

0.3

0.2

0.2

0.1

0.1

PGK

TUB

EXP

TIP41

PP2A

CYP

40S

EF-1α

ACT

F-box

PP2A

40S

UBQ

GAPDH

PGK

18S

EF-1α

ACT

TUB

GAPDH TUA

0 EXP

UBQ

TUB

GAPDH

18S

EF-1α

TIP41

40S

PP2A

TUA

CYP

0.4

0.25

TIP41 F-box

F-box

ACT PGK

UV-C

Salinity

0.225

EXP

Cold 0.9

18S

CYP

1

0.9

0 UBQ

Average expression stability (M)

0 TUA

0.35

0.2 0.3

0.175 0.25

0.15

0.2

0.125 0.1

0.15

0.075 0.1

0.05 0.05

0.025 0 TUB

UBQ

GAPDH

EXP

TUA

PP2A

TIP41

EF-1α

PGK

CYP

18S

ACT

0.9

40S F-box

Average expression stability (M)

ABA 0.8

0 UBQ

18S

EF-1α

EXP

40S

CYP

GAPDH

PGK

TUA

F-box

TIP41

ACT PP2A

MeJA

0.55 0.5

0.7

0.45

0.6

0.4

0.5

0.35 0.3

0.4

0.25

0.3

0.2

0.2

0.15 0.1

0.1 0 EXP

TUB

0.6

0.05 UBQ

TIP41

TUB

GAPDH

PGK

<::::: Least stable genes

18S

EF-1α

F-box

TUA

Most stable genes ::::>

CYP

ACT

40S PP2A

0 UBQ

TUA

EXP

TUB

PGK

18S

GAPDH

<::::: Least stable genes

EF-1α

F-box

TIP41

40S

ACT

PP2A CYP

Most stable genes ::::>

Fig. 2. Average expression stability values (M) of the 14 candidate reference genes using geNorm software. Expression stability was evaluated in samples from Hedera helix L. submitted to total samples, total tissues samples, total abiotic stresses samples, cold stress, heat stress, drought stress, salinity stress, UV-C radiation, ABA and MeJA treatment. The most stable reference genes were measured during stepwise exclusion of the least stable reference genes. The least stable genes are on the left and the most stable genes on the right.

H.-p. Sun et al. / Plant Physiology and Biochemistry 108 (2016) 286e294

Pairwise variation (V)

0.15

291

Tissues Total Heat Salinity ABA

Abiotic stress Cold Drought UV-C MeJA

0.10

0.05

0.00 V2/3

V3/4

V4/5

V5/6

V6/7

V7/8

V8/9

V9/10

V10/11

V11/12

V12/13

V13/14

Fig. 3. Pairwise variation (V) analysis of the fourteen candidate reference genes. The pairwise variation (Vn/Vnþ1) was analysed between the normalization factors Vn and Vnþ1 by the geNorm software to determine the optimal number of reference genes required for RT-qPCR data normalization.

Table 2 Ranking of 14 candidate reference genes under different experimental conditions in order of their expression stability calculated by NormFinder. Rank Total

Tissuesa

Abiotic stressesb Cold

Heat

Drought

Salinity

UV-C

ABA

MeJA

1 2 3 4 5 6 7 8 9 10 11 12 13 14

F-box(0.165) ACT(0.167) PP2A(0.222) EF-1a(0.232) TIP41(0.235) TUA (0.277) 40S(0.336) GAPDH(0.398) CYP(0.444) PGK(0.457) 18S(0.492) TUB(0.505) UBQ(0.596) EXP(0.677)

F-box(0.162) ACT(0.194) EF-1a(0.196) 40S(0.286) CYP(0.352) TIP41(0.356) GAPDH(0.374) PP2A(0.375) TUA(0.415) TUB(0.458) 18S(0.511) PGK(0.531) EXP(0.627) UBQ(0.713)

TUA(0.041) GAPDH(0.099) EF-1a(0.099) F-box(0.104) ACT(0.146) CYP(0.248) 40S(0.277) TIP41(0.370) PP2A(0.409) EXP(0.546) PGK (0.670) TUB(0.685) 18S(0.758) UBQ(0.788)

TIP41(0.180) F-box(0.190) ACT(0.212) EF-1a(0.237) TUB(0.243) EXP(0.248) 40S(0.285) 18S(0.315) PP2A(0.332) PGK (0.366) CYP(0.386) UBQ(0.397) GAPDH(0.409) TUA(0.428)

F-box(0.070) 40S(0.113) CYP(0.158) 18S(0.161) ACT(0.166) EF-1a(0.179) TUA(0.202) PP2A(0.205) TIP41(0.205) PGK (0.241) UBQ(0.219) GAPDH(0.245) EXP(0.294) TUB(0.410)

F-box(0.065) PP2A(0.074) ACT (0.088) TIP41(0.090) GAPDH(0.108) TUA(0.108) PGK(0.113) EF-1a(0.142) 18S(0.174) EXP(0.174) CYP(0.201) 40S(0.221) TUB(0.271) UBQ(0.284)

TUA(0.159) 40S(0.173) CYP(0.181) PP2A(0.337) ACT(0.351) TIP41(0.355) F-box(0.358) PGK(0.364) 18S(0.370) EF-1a(0.376) GAPDH(0.385) TUB(0.390) UBQ(0.412) EXP(0.562)

F-box(0.111) GAPDH(0.141) TIP41(0.151) ACT(0.158) CYP(0.180) EF-1a(0.185) PGK(0.215) 18S(0.216) PP2A(0.217) 40S(0.225) TUB(0.240) UBQ(0.330) EXP(0.332) TUA(0.351)

a b

F-box(0.222) EF-1a(0.254) ACT(0.255) TIP41(0.376) 40S(0.381) CYP(0.419) PP2A(0.454) 18S(0.466) GAPDH(0.473) TUA (0.516) PGK(0.554) TUB(0.585) UBQ(0.778) EXP(0.875)

F-box(0.142) PGK (0.216) ACT(0.257) TIP41(0.286) CYP(0.339) PP2A(0.352) EF-1a(0.352) 18S (0.387) TUA(0.392) GAPDH(0.416) 40S(0.587) TUB(0.604) UBQ(0.746) EXP(0.795)

Tissues analyses include all tissues together. Abiotic stresses analyses include all stress abiotic together.

Raw Ct value can directly reflect the reference gene expression level in test samples to evaluate the reference genes by RT-qPCR (Bustin, 2009; Yan et al., 2011). Generally, a higher Ct value indicates a lower expression level of the reference gene in the test sample. Therefore, the raw Ct value can provide basic data support to evaluate candidate reference genes (Chen et al., 2015). Within an appropriate range of Ct values, the candidate reference gene can be further analysed by software and algorithms. The raw Ct values in our results also indicated that none of the selected reference genes

demonstrated constant expression in different samples. The selection and evaluation of the suitable reference gene(s) by software and algorithms were extremely important for gene expression normalization under different treatments in H. helix. Next, we chose geNorm and NormFinder for stability evaluation of the candidate reference genes. These two software programs are the most widely applied algorithms in the study of reference gene selection in several plants, such as lettuce (Borowski et al., 2014), banana (Chen et al., 2011), switchgrass (Huang et al., 2014),

Table 3 Consensus of stability ranking of the reference gene estimated by geNorm and NormFinder.

*

Different sample sets

The five most stable reference gene

Most stable combination of two genes

The least stable reference gene

Total Tissues Abiotic stresses Cold Heat Drought Salinity UV-C ABA MeJA

ACT, F-box, 40S, PP2A*(EF-1a)**, TUA*(TIP41)** F-box, TIP41, ACT, PP2A, EF-1a ACT, F-box, 40S, EF-1a, PP2A* (CYP) ** PGK, ACT, F-box, CYP, TUA* (TIP41)** GAPDH, TUA, F-box, ACT, EF-1a TIP41, F-box, TUB, ACT, EXP* (EF-1a)** F-box, 40S, ACT, 18S, CYP ACT, PP2A, TIP41, F-box, TUA* (GAPDH)** 40S, PP2A, ACT, CYP, TUA, CYP, ACT, TIP41, 40S*(F-box)**, PP2A*(GAPDH)**

F-box þ ACT* (EF-1a)** F-box þ TIP41* (ACT) ** F-box þ ACT PGK þ ACT* (F-box)** GAPDH þ TUA TIP41þF-box F-box þ 40S PP2A þ ACT* (F-box)** 40S þ PP2A* (TUA)** CYP þ PP2A*, F-box þ GAPDH**

EXP EXP UBQ EXP UBQ TUA TUB UBQ EXP UBQ*, TUA**

or

**

indicated that stability ranking of the reference gene was estimated by geNorm and NormFinder, respectively.

292

H.-p. Sun et al. / Plant Physiology and Biochemistry 108 (2016) 286e294

Fig. 4. Relative expression of HhSE and Hhb-AS using selected reference genes including the most or the least stable reference genes for normalization under different tissues and cold treatment experimental sets. a HhSE expression of different tissues; b Hhb-AS expression of different tissues; c HhSE expression of leaves under cold treatment after 0 d, 5 d, 15 d and 30 d; d Hhb-AS HhSE expression of leaves under cold treatment after 0 d, 5 d, 15 d and 30 d. The error bars represent standard errors. T-test were generated by ANOVA among relative expression levels in same sample and denoted as follows: *p < 0.05, **p < 0.01.

creeping bentgrass (Chen et al., 2015), Caragana korshinskii (Yang et al., 2014), A. belladonna (Li et al., 2014), soybean (Ma et al., 2013) and garlic (Liu et al., 2015). Other reports simultaneously evaluated the use of the BestKeeper software (Pollier et al., 2014; Li et al., 2014; Chen et al., 2015; Liu et al., 2015). Although BestKeeper is a completely different algorithm from geNorm and NormFinder, this software can simultaneously evaluate only less than 10 reference genes (Pfaffl, 2001; Pfaffl et al., 2004). We selected 14 candidate reference genes in our experimental sets, however, so their results were beyond the maximum evaluation scope of BestKeeper. Thus, we opted not use BestKeeper in this study. A large number of reports on reference gene selection and evaluation in different plants or animal varieties confirmed that an ideal reference gene was non-existent. Similarly, our results showed that reference genes with stable expressions differed for different tissues and treatments. First, the geNorm analysis results showed that all of the M values of the 14 candidate reference genes did not exceed the 1.5 threshold in all analysis sets, regardless of the total analysis sets or the single analysis sets. However, the results identified the eight most stable combinations among these 10 analysis sets, wherein only the sets of total samples and total abiotic stress treatments, the total tissues samples and drought stress obtained the same most stable combinations, respectively. Second, the NormFinder results further verified the same points as geNorm. Based on the stability value sequence, three genes ranked first in all analysis sets: F-box, TUA and TIP41. Different studies reported varied conclusions on the consistency of the geNorm and NormFinder results. In a comprehensive comparison of geNorm and NormFinder analysis results, a part of the analysis sets had the same top five most stable genes, whose stability values were analysed by both geNorm and NormFinder. A number of sets indicated one or two different gene(s) among the

top five stable genes (Chen et al., 2011; Yang et al., 2014). However, even the sets with the same top five stable genes were not completely identical in the analysis result orders for geNorm and NormFinder. Furthermore, the use of multiple reference genes in combinations was more accurate for normalizing the expression levels of target genes. In this study, geNorm and NormFinder showed that a two-gene combination was the minimum number under different treatments in H. helix (Vandesompele et al., 2002; Claus et al., 2004). The most stable combination calculated by these two algorithms demonstrated the same results as a single stable gene. As indicated by the abovementioned analysis, the stability values were not identical as obtained by different software programs with various algorithms to evaluate H. helix reference genes. A validation experiment on reference gene selection and evaluation is necessary (Chervoneva et al., 2010). We choose HhSE and Hhb-AS as target genes because HhSE and Hhb-AS are key regulatory enzymes in the biosynthesis pathway of triterpenoid saponins; their expression levels were positively correlated with the triterpenoid saponin content in H. helix. Simultaneously, the selection of different tissues and cold stress as experimental sets was based on the regularity derived in our previous studies. The triterpenoid saponin content was highest in the leaf compared to different H. helix tissues, followed by the stem and the root. In the cold stress treatment, the triterpenoid saponin content presented a downward trend. As shown in Fig. 4aeb, the expression trends of HhSE and Hhb-AS in different tissues were identical with our previous results. Evidently, small changes were noted when the most stable reference genes were selected to normalize their relative expression levels, regardless of whether a single gene or the best combination of two genes was used. By contrast, large changes and standard deviations were noted when the most unstable reference genes

H.-p. Sun et al. / Plant Physiology and Biochemistry 108 (2016) 286e294

were used for normalization. The relative expression patterns of HhSE and Hhb-AS in Fig. 4ced were increasingly apparent under cold stress, with consistent results when the most stable reference genes were used for normalization. However, the correct patterns were not obtained when the least stable reference genes were used. These validation results confirmed the applicability and correctness of the reference genes selected and evaluated in H. helix. These results suggested that the selection and evaluation of stable reference genes represent a crucial issue for the proper normalization of the RT-qPCR data. In conclusion, we selected and evaluated 14 candidate reference genes to normalize gene expression analysis in five tissues of H. helix and under seven abiotic stress conditions. To the best of our knowledge, this study is the first to select and evaluate reference genes in H. helix for the normalization of gene expression analysis through RT-qPCR. Based on our results, we recommend that suitable reference genes in H. helix should be selected for normalization relative to different experimental sets. F-box yielded a higher stability than the other reference genes. It ranked at the top three suitable genes under all experimental setups except for ABA treatment. EXP and UBQ were the least stable reference genes in H. helix. In addition, the most stable combination suggested by our results was that F-box and ACT were associated with different tissues and total abiotic stresses; F-box and TIP41, 40S, PP2A, GAPDH were associated with drought, salinity, UV-C and MeJA treatments, respectively; PGK and ACT were associated with cold stress; GAPDH and TUA were associated with heat stress; 40S and PP2A were associated with ABA treatment. The HhSE and Hhb-AS expression analysis confirmed the importance of reference gene validation in obtaining accurate RT-qPCR results. These selected stable reference genes collectively supply an important foundation for using RTqPCR for an accurate and far-reaching gene expression analysis in Hedera helix L. Contribution Huapeng Sun and Xiaohong Zhong conceived and designed the experiments. Huapeng Sun, Fang Li performed the experiments. Huapeng Sun, Fang Li and Qinmei Ruan analysed the data and Huapeng Sun wrote the manuscript guided by XZ. All authors read and approved the final manuscript. Conflict of interest The authors declare no conflicts of interest in the submission of this manuscript. Acknowledgements This study was financed by the Education Department of Hunan Province (Project No. 15A089) and the Graduate Innovative Projects in Hunan Province (Project No. CX2014B290). We would like to thank the National Centre for Citrus Improvement Changsha for providing the experimental platform. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.plaphy.2016.07.022. References Borges, A.F., Fonseca, C., Ferreira, R.B., Lourenco, A.M., Monteiro, S., 2014. Reference gene validation for quantitative RT-PCR during biotic and abiotic stresses in Vitis vinifera. Plos One 9, e111399. http://dx.doi.org/10.1371/

293

journal.pone.0111399. Borowski, J.M., Galli, V., Messias, R.S., Perin, E.C., Buss, J.H., Silva, S.D.A., Rombaldi, C.V., 2014. Selection of candidate reference genes for real-time PCR studies in lettuce under abiotic stresses. Planta 239, 1187e1200. Bustin, S.A., 2002. Quantification of mRNA using real-time reverse transcription PCR (RT-PCR): trends and problems. J. Mol. Endocrinol. 29, 4021e4022. Bustin, S.A., 2009. Why the need for qPCR publication guidelines-The case for MIQE. Methods 50, 217e226. Chen, L., Zhong, H.Y., Kuang, J.F., Li, J.G., Lu, W.J., Chen, J.Y., 2011. Validation of reference genes for RT-qPCR studies of gene expression in banana fruit under different experimental conditions. Planta 234, 77e390. Chen, Y., Hu, B., Tan, Z., Liu, J., Yang, Z., Li, Z., Huan, B., 2015. Selection of reference genes for quantitative real-time PCR normalization in creeping bentgrass involved in four abiotic stresses. Plant Cell Rep. 34, 1825e1834. Chervoneva, I., Li, Y., Schulz, S., Croker, S., Wilson, C., Waldman, S.A., Hyslop, T., 2010. Selection of optimal reference genes for normalization in quantitative RT-PCR. BMC Bioinforma. 11, 1e15. Claus, L.A., Jens, L.J., Torben, F.R., 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, 5245e5250. Colebatch, G., Trevaskis, B., Udvardi, M., 2002. Functional genomics: tools of the trade. New Phytol. 153, 27e36. Cwientzek, U., Ottillinger, B., Arenberger, P., 2011. Acute bronchitis therapy with ivy leaves extracts in a two-arm study. A double-blind, randomised study vs. an other ivy leaves extract. Phytomedicine 18, 1105e1109. Derveaux, S., Vandesompele, J., Hellemans, J., 2010. How to do successful gene expression analysis using real-time PCR. Methods 50, 227e230. Fazio, S., Pouso, J., Dolinsky, D., Fernandez, A., Hernandez, M., Clavier, G., Hecker, M., 2009. Tolerance, safety and efficacy of Hedera helix extract in inflammatory bronchial diseases under clinical practice conditions: a prospective, open, multicentre postmarketing study in 9657 patients. Phytomedicine 16, 17e24. Gachon, C., Mingam, A., Charrier, B., 2004. Real-time PCR: what relevance to plant studies? J. Exp. Bot. 55, 1445e1454. Gepdiremen, A., Mshvildadze, V., Süleyman, H., Elias, R., 2005. Acute antiinflammatory activity of four saponins isolated from ivy: alpha-hederin, hederasaponin-C, hederacolchiside-E and hederacolchiside-F in carrageenaninduced rat paw edema. Phytomedicine 12, 440e444. Han, J.Y., In, J.G., Kwon, Y.S., Choi, Y.E., 2010. Regulation of ginsenoside and phytosterol biosynthesis by RNA interferences of squalene epoxidase gene in Panax ginseng. Phytochemistry 71, 36e46. Haralampidis, K., Trojanowska, M., Osbourn, A.E., 2002. Biosynthesis of triterpenoid saponins in plants. Adv. Biochem. Eng. Biot. 75, 31e49. €lp, A., 2003. Efficacy of dry extract of ivy leaves in Hofmann, D., Hecker, M., Vo children with bronchial asthmaea review of randomized controlled trials. Phytomedicine 10, 213e220. Holzinger, F., Chenot, J.F., 2011. Systematic review of clinical trials assessing the effectiveness of ivy leaf (Hedera helix) for acute upper respiratory tract infections. Evid-based. Compl. Alt. 1, 1e9. Huang, L., Yan, H., Jiang, X., Zhang, X., Zhang, Y., Huang, X., Zhang, Y., Miao, J., Xu, B., Frazier, T., 2014. Evaluation of candidate reference genes for normalization of quantitative RT-PCR in switchgrass under various abiotic stress conditions. Bioenergy. Res. 7, 1201e1211. Ilhami, G.I., Vakhtang, M., Ahan, G., Riad, E., 2004. Antioxidant activity of saponins isolated from ivy: alpha-hederin, hederasaponin-C, hederacolchiside-E and hederacolchiside-F. Planta. Med. 70, 561e563. Jacinta, G., Nicholas, E., Allen, V.D., Eduardo, B., 2014. Selection and validation of reference genes for gene expression analysis in switchgrass (Panicum virgatum) using quantitative real-time RT-PCR. Plos One 9, e91474. Khdair, A., Mohammad, M.K., Tawaha, K., Al-Hamarsheh, E., AlKhatib, H.S., AlKhalidi, B., Bustanji, Y., Najjar, S., Hudaib, M., 2010. A validated RP HPLC-PAD method for the determination of hederacoside C in Ivy-Thyme cough syrup. Int. J. Anal. Chem. http://dx.doi.org/10.1155/2010/478143. Kozera, B., Rapacz, M., 2013. Reference genes in real-time PCR. J. Appl. Genet. 54, 391e406. Kushiro, T., Shibuya, M., Ebizuka, Y., 1998. Beta-amyrin synthaseecloning of oxidosqualene cyclase that catalyzes the formation of the most popular triterpene among higher plants. Eur. J. Biochem. 256, 38e244. Landgrebe, H., Matusch, R., Marburg, F.R., Hecker, M., 1999. Effectiveness and use of an old medicinal plant. Pharm. Ztg. N. R. 35, 11e15. Li, J., Chen, M., Qiu, F., Qin, B., Liu, W., Wu, N., Lan, X., Wang, Q., Liao, Z., Tang, K., 2014. Reference gene selection for gene expression studies using quantitative real-time PCR normalization in Atropa belladonna. Plant. Mol. Biol. Rep. 32, 1002e1014. Liu, C., Jeng, W., Chang, W., Shih, M., Ko, T., Wang, A., 2014a. Structural insights into the catalytic mechanism of human squalene synthase. Acta Crystallogr. 70, 231e241. Liu, J., Wang, Q., Sun, M., Zhu, L., Yang, M., Zhao, Y., 2014b. Selection of reference genes for quantitative real-time PCR normalization in Panax ginseng at different stages of growth and in different organs. Plos One 9, e112177. Liu, M., Wu, Z., Jiang, F., 2015. Selection and validation of garlic reference genes for quantitative real-time PCR normalization. Plant. Cell. Tiss. Org. 122, 435e444. Luo, H., Chao, S., Sun, Y., Wu, Q., Ying, L., Song, J., Niu, Y., Cheng, X., Xu, H., Li, C., 2011. Analysis of the transcriptome of Panax notoginseng root uncovers putative triterpene saponin-biosynthetic genes and genetic markers. BMC. Genomics 12,

294

H.-p. Sun et al. / Plant Physiology and Biochemistry 108 (2016) 286e294

1e15. Lutsenko, Y., Bylka, W., Matlawska, I., Darmohray, R., 2010. Hedera helix as a medicinal plant. Herba. Pol. 56, 83e96. Ma, S., Niu, H., Liu, C., Zhang, J., Hou, C., Wang, D., 2013. Expression stabilities of candidate reference genes for RT-qPCR under different stress conditions in soybean. Plos One 8, 108e110. Marum, L., Miguel, A., Ricardo, C.P., Miguel, C., 2012. Reference gene selection for quantitative real-time PCR normalization in Quercus suber. Plos One 7, e35113. Mendel, M., Chłopecka, M., Dziekan, N., Wiechetek, M., 2011. The effect of the whole extract of common ivy (Hedera helix) leaves and selected active substances on the motoric activity of rat isolated stomach strips. J. Ethnopharmacol. 134, 796e802. Mendel, M., Chłopecka, M., Dziekan, N., Wiechetek, M., 2013. Participation of extracellular calcium in a-hederin-induced contractions of rat isolated stomach strips. J. Ethnopharmacol. 146, 423e426. Pfaffl, M.W., 2001. A new mathematical model for relative quantification in realtime RT-PCR. Nucleic. Acids. Res. 29, e45. Pfaffl, M.W., Tichopad, A., Prgomet, C., Neuvian, T.P., 2004. Determination of stable housekeeping genes, differentially regulated target genes and sample integrity: BestKeepereExcel-based tool using pair-wise correlations. Biotechnol. Lett. 26, 509e515. Pollier, J., Bossche, R.V., Rischer, H., Goossens, A., 2014. Selection and validation of reference genes for transcript normalization in gene expression studies in Catharanthus roseus. Plant. Physiol. Bioch 83C, 20e25. Reid, K.E., Olsson, N., Schlosser, J., Peng, F., Lund, S.T., 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. Stauss-Grabo, M., Atiye, S., Warnke, A., Wedemeyer, R., Donath, F., Blume, H., 2011. Observational study on the tolerability and safety of film-coated tablets containing ivy extract (Prospan® Cough Tablets) in the treatment of colds accompanied by coughing. Phytomedicine 18, 433e436. Vandesompele, J., Preter, K.D., Pattyn, F., Poppe, B., Roy, N.V., Paepe, A.D., Speleman, F., 2002. Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome. Biol. 3 research0034.1e0034.11. Wu, Q., Ma, X., Zhang, K., Feng, X., 2015. Identification of reference genes for tissuespecific gene expression in Panax notoginseng using quantitative real-time PCR. Biotechnol. Lett. 37, 197e204. Yan, J., Yuan, F., Long, G., Qin, L., Deng, Z., 2011. Selection of reference genes for quantitative real-time RT-PCR analysis in citrus. Mol. Biol. Rep. 39, 1831e1838. Yang, Q., Yin, J., Li, G., Qi, L., Yang, F., Wang, R., Li, G., 2014. Reference gene selection for qRT-PCR in Caragana korshinskii Kom. under different stress conditions. Mol. Biol. Rep. 41, 2325e2334. Ye, Y., Wang, R., Jin, L., Shen, J., Li, X., Yang, T., Zhou, M., Yang, Z., Chen, Y., 2014. Molecular cloning and differential expression analysis of a squalene synthase gene from Dioscorea zingiberensis, an important pharmaceutical plant. Mol. Biol. Rep. 41, 6097e6104. Yuan, J.S., Galbraith, D.W., Dai, S.Y., Griffin, P., Stewart, C.N., 2008. Plant systems biology comes of age. Trends. Plant. Sci. 13, 165e171.