Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Selection of appropriate reference genes for quantitative real-time reverse transcription PCR in Betula platyphylla under salt and osmotic stress conditions

Selection of appropriate reference genes for quantitative real-time reverse transcription PCR in... Selecting appropriate reference genes is vital to normalize gene expression analysis in birch (Betula platyphylla) under different abiotic stress conditions using quantitative real- OPENACCESS time reverse transcription PCR (qRT-PCR). In this study, 11 candidate birch reference genes (ACT, TUA, TUB, TEF, 18S rRNA, EF1α, GAPDH, UBC, YLS8, SAND, and CDPK) Citation: Li Z, Lu H, He Z, Wang C, Wang Y, Ji X (2019) Selection of appropriate reference genes for were selected to evaluate the stability of their expression in different tissues and under dif- quantitative real-time reverse transcription PCR in ferent abiotic stress conditions. Three statistical algorithms (GeNorm, NormFinder, and Betula platyphylla under salt and osmotic stress BestKeeper) were used to analyze the stability of the 11 candidate reference genes to iden- conditions. PLoS ONE 14(12): e0225926. https:// tify the most appropriate one. The results indicated that EF-1α was the most stable refer- doi.org/10.1371/journal.pone.0225926 ence gene in different birch tissues, ACT was the most stable reference gene for normal Editor: Mayank Gururani, United Arab Emirates conditions, ACT and TEF were the most stable reference genes for salt stress treatment, University, UNITED ARAB EMIRATES TUB was the most stable reference gene for osmotic stress treatment, and ACT was the Received: September 9, 2019 most appropriate choice in all samples of birch. In conclusion, the most appropriate refer- Accepted: November 15, 2019 ence genes varied among different experimental conditions. However, in this study, ACT Published: December 3, 2019 was the optimum reference gene in all experimental groups, except in the different tissues Peer Review History: PLOS recognizes the group. GAPDH was the least stable candidate reference gene in all experimental conditions. benefits of transparency in the peer review In addition, three stress-induced genes (BpGRAS1, BpGRAS16, and BpGRAS19) were process; therefore, we enable the publication of chosen to verify the stability of the selected reference genes in different tissues and under all of the content of peer review and author salt stress. This study laid the foundation for the selection of appropriate reference gene(s) responses alongside final, published articles. The editorial history of this article is available here: for future gene expression pattern studies in birch. https://doi.org/10.1371/journal.pone.0225926 Copyright:© 2019 Li et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and Introduction source are credited. Quantitative real-time reverse transcription polymerase chain reaction (qRT-PCR) has Data Availability Statement: All relevant data are become a common technique to detect gene expression levels in the field of molecular biology within the manuscript and its Supporting Information files. because of its speed, high efficiency, labor-saving, and sensitivity. It has been widely used in PLOS ONE | https://doi.org/10.1371/journal.pone.0225926 December 3, 2019 1 / 13 Selection of appropriate reference genes in birch Funding: This work was supported by the National basic biological research [1], food spoilage detection [2], environmental monitoring [3], and Natural Science Foundation of China (No. diagnosis and treatment of disease [4]. QRT-PCR has strict requirements on the purity and 31500535), and The Heilongjiang Province Science quality of RNA, reverse transcription efficiency, the specificity of the primers, PCR amplifica- Foundation (No. QC2018017). tion efficiency, and the selection of internal reference genes [5]. Reference genes can have a Competing interests: The authors have declared large influence on the results of relative expression level determination. Thus, selecting ideal that no competing interests exist. reference genes is vital to obtain reliable results. One or more internal reference genes are typi- cally required to normalize qRT-PCR data to obtain more accurate results. Over the course of an experiment, a reliable internal control should show a minimal change in expression, whereas the expression of a gene of interest may change greatly. Housekeeping genes are often selected because they encode proteins essential to cell viability and therefore are assumed to be expressed similarly in different cell types [6]. Commonly used internal con- trols include actin (ACT), beta-tubulin (TUB), elongation factor 1 alpha (EF1-α), and glyceral- dehyde-3-phosphate dehydrogenase (GAPDH) [7–8]. However, increasing numbers of reports have confirmed that the expression of housekeeping genes can vary in different tissues and may be affected by the experimental conditions. For instance, TUB is a wildly used reference gene; a study showed that it was unsuitable as a reference gene in different tissues of Pennise- tum glaucum [9]. The TUA (α-tubulin) gene was selected as reference gene in Salicornia euro- paea under salt stress [10–11]; whereas it was not considered as an ideal reference gene in S. europaea under drought treatment [12]. ACT was selected as a reference gene in wheat and Kosteletzkya virginica under salt stress, but was not suitable as a reference gene in papaya under many experimental conditions or in Salicornia europaea and cucumber under salt stress [12–16]. Therefore, selecting the appropriate reference genes according to different tissues and different experimental conditions is essential. Studies on the evaluation and validation of reference genes have been conducted in many plant species, especially in herbaceous plants, such as Arabidopsis thaliana [17], Glycine max [18]. Similar studies have been carried out in certain woody plants, such as Morus alba L. [19], Carex rigescens [20]. However, the systematic evaluation and validation of appropriate refer- ence genes in birch (Betula platyphylla) under different abiotic stress conditions has not been reported. Betula platyphylla, which is a species of deciduous hardwood, is widely distributed in the mid-high mountains of warm temperate regions in the world, including northern China, Rus- sian Far East, Siberia, Mongolia, Northern Korea, and Japan [21]. With the continuous expan- sion of drought and saline areas worldwide, it is of great significance to cultivate new salt tolerant and drought resistant varieties of birch. Therefore, selection of internal reference genes in birch with stable expression under different abiotic stress conditions plays a key role in studies on stress-related gene expression. The analysis results of qRT-PCR were affected to varying degrees of influence in reference genes choosing. Some studies show how to choose genes as reference genes to normalized expression data. [22]. Therefore, in this study, eleven candidate reference genes (ACT, TUA, TUB, TEF (Translation elongation factor)), 18S rRNA (18S ribosomal RNA), EF1α, GAPDH, UBC (Ubiquitin-conjugating enzyme E2), YLS8 (Thioredoxin-like protein YLS8), SAND (S- adenosyl methionine decarboxylase), and CDPK (CDPK-related kinase) were selected to eval- uate their expression stability in different tissues and under abiotic stress conditions in birch. Three different statistical tools (GeNorm, NormFinder, and BestKeeper) were used to analyze the stability of 11 candidate reference genes and identify the most appropriate reference gene. Our results will facilitate of appropriate reference genes selection for the development of gene expression pattern studies in birch. PLOS ONE | https://doi.org/10.1371/journal.pone.0225926 December 3, 2019 2 / 13 Selection of appropriate reference genes in birch Materials and methods Plant material and stress treatments In this experiment, open pollinated northeast white birch (B. platyphylla) seeds from the North- east Forestry University were planted in plastic pots with mixture of turf peat and sand (v/v 3:1), -2 -1 in a growth house where the conditions kept at 70–75% relative humidity, 400 μmol�m s light intensity, a 16 h light/8 h dark photocycle at 25±2˚C. The 8-week-old birch plants grown in soil were watered with a solution of 2 L 200 mM NaCl or 2 L 300 mM mannitol (distilled water served as the control) for 3, 6, 12, 24, and 48 h, respectively. The root, stem, and leaf tissues from 6 birch seedlings were collected after each treatment. Three independent replications were performed. All the samples were placed at −80˚C after freezing in liquid nitrogen for further study. Total RNA isolation and cDNA synthesis The total RNA of all samples was isolated using a Universal Plant Total RNA Extraction Kit (Bioteke, Beijing, China) following the manufacturer’s instructions. RNA integrity was checked using 1.5% agarose gel electrophoresis, and the RNA concentration and purity were determined using a NanoDrop 2000 Spectrophotometer (NanoDrop, Thermo Scientific, Wal- tham, MA, USA). All the RNA samples with A260/A280 ratios = 1.9–2.1 and A260/A230 ratios > 2.0 were used for further study. Then RNA (1 μg) was reversely transcribed into cDNA using a TransScript One-Step gDNA Removal and cDNA Synthesis SuperMIX (Trans- Gen Biotech, Beijing, China) following the manufacturer’s instructions. The synthesized cDNA was stored at −20˚C for further study. Candidate reference gene selection and primer design Eleven candidate reference genes that are commonly used in qRT-PCR were chosen for this study ACT, TUA, TUB, TEF, 18S rRNA, EF1α, GAPDH, UBC, YLS8, SAND, and CDPK [22– 23]. The GenBank accession numbers for the 11 genes, which were obtained from the birch genome data (unpublished), are listed in Table 1. Specific primers for the reference genes were designed using Primer premier 5.0 (Premier Biosoft International, Palo Alto, CA, USA). The primer sequences for qRT-PCR are listed in Table 1. Quantitative real time PCR A TransStart Top Green qPCR SuperMix kit (TransGen Biotech, Beijing, China) was used for qRT-PCR. All qRT-PCR reactions were carried out on Qtower G (Analytik, Jena, Germany). The qRT-PCR reaction system included: 10 μL of 2 × TransStart Top Green Realtime qPCR SuperMix, 0.4 μL upstream primers, and 0.4 μL downstream primers (10 μM), 2 μL of cDNA templates (100 ng), and 7.2 μL of RNAase-free ddH O with a volume of 20 μL. The qRT-PCR reaction conditions were as follows: 94˚C for 30 s; 45 cycles at 94˚C for 12 s, 60˚C for 30 s, and 72˚C for 40 s; and 79˚C for 1 s for plate reading. The melting curve was constructed to verify the specificity of each sequence-specific primer. The PCR amplification efficiency (E) was calculated as E = −1 + 10 (−1/ slope) [24], where the slope was derived from a standard curve generated by 10-fold serial dilu- tions of the mixed cDNA for each primer pair. Three independent replications were performed. Data analysis SigmaPlot 10 software (Systat Software, Inc., San Jose, CA, USA) was used to show the cycle threshold (Ct) value distribution of all samples, which can reflect clearly the average expression level of the genes in the samples. Threshold for the Ct values is the machine setting. The aver- age Ct value was calculated using three biological replications. GeNorm [25] and NormFinder PLOS ONE | https://doi.org/10.1371/journal.pone.0225926 December 3, 2019 3 / 13 Selection of appropriate reference genes in birch Table 1. Primer sequences and amplification parameters of 11 candidate reference genes from birch. Gene symbol Gene name Accession Primers Tm E Amplicon Mean Ct SD CV (%) 0 0 no. (5 -3 ) length (bp) 18S rRNA 18S ribosomal RNA MK388236 F: AACGAACGAGACCTCAGCCT 61.18 1.97 171 17.39 1.36 14.01 R: ACTCGTTGAATACATCAGTG 53.09 GAPDH Glyceraldehyde-3-phosphate dehydrogenase MK388226 F: AAGCTCAATGGCATTGCACT 58.74 1.89 205 26.87 2.33 15.64 R: TGGAAGAAACATCAGTGCAC 56.26 ACT Actin MK388227 F: TGAGAAGAGCTATGAGTTGC 54.89 1.98 204 22.49 1.03 8.26 R: GTAGATCCACCACTAAGCAC 55.27 TUA Alpha-tubulin MK388228 F: ATATCATCCTTGACAACTTC 50.18 1.94 210 22.80 1.00 7.92 R: GATGCCACATTTGAAGCCAG 57.71 TUB beta-tubulin MK388229 F: GTTAGCGAGCAGTTTACAGC 57.21 2.01 195 23.22 1.05 8.10 R: ACCAACAACCTCTTCTTCTT 54.39 UBC Ubiquitin-conjugating enzyme E2 MK388230 F: CAAATGACAGTCCCTATGCT 55.12 1.99 219 26.05 1.81 12.48 R: GGTCAGTAAGCAATGAACAT 53.58 EF-1α Elongation factor 1-alpha MK388231 F: GACAACGTTGGCTTCAACGT 59.62 1.92 203 23.09 1.96 15.27 R: GCAAACTTCACAGCAATGTG 56.43 TEF Translation elongation factor MK388232 F: GTACAATGATGAGAACACTG 51.64 1.96 201 22.28 1.41 11.38 R: TCCATGTCATAGCACTCATC 54.64 SAND Sadenosyl methionine decarboxylase MK388233 F: GCATCTAGGACAACCTAGAG 54.38 1.86 213 26.19 1.66 11.40 R: CCACTATCATGCATAGAAGC 53.65 YLS8 Thioredoxin-like protein YLS8 MK388234 F: GCATCTGTTGCTGAGACAAT 56.42 1.90 216 23.47 1.53 11.71 R: CTCCACAATATCAATGAACT 50.48 CDPK CDPK-related kinase MK388235 F: GAAGATGAGCTCATCTACCT 53.66 1.93 212 29.80 2.06 12.45 R: TAAGTACTGATTGCAGCAGC 55.85 Tm, Melting Temperature; E, PCR amplification efficiency; Ct, cycle threshold; CV, coefficient of variation; SD, standard deviation https://doi.org/10.1371/journal.pone.0225926.t001 [26] software were used to analyze the stabilities of the candidate reference genes’ expression. The mean Ct value must be transformed into relative expression level based on the formula -ΔCt -ΔCt E (-ΔCt = Ct value of each sample-the minimum Ct value) [27]. The E value can be ana- lyzed by GeNorm and NormFinder to obtain the candidate reference gene’s expression stabil- ity. GeNorm can determine the stability of each candidate reference gene through calculating the M value according to its expression. When the M value is less than 1.5, the gene can be used as an appropriate reference gene. Moreover, the lower the M value, the more stable the expression of the candidate reference gene. NormFinder can rank the stability of candidate reference genes expression. Moreover, it can identify the best reference gene among all candi- date reference genes. The BestKeeper [28] program analyzed the pairing correlation under a given set of experimental conditions using the Ct values of the reference genes in each group. According to the standard deviation (SD) and coefficient of correlation of the candidate genes R value, the expression stability of each reference gene was evaluated in the BestKeeper pro- gram. A reference gene with a lower SD value and an R value closer to 1 would have stable expression. Correlation analysis (R ) using SPSS (19.0) between M values vs SV values, SV value vs SD values, M values vs SD values for determine its data robustness. A simple T test using SPSS to illustrate the significance of data differences. Differences were considered to be significant if P < 0.05. represented 0.01 < P < 0.05. Validation of reference genes The identified optimum reference genes and the least stable reference genes in salt or in differ- ent tissues were used to normalize the relative expression levels of BpGRAS1 (GenBank PLOS ONE | https://doi.org/10.1371/journal.pone.0225926 December 3, 2019 4 / 13 Selection of appropriate reference genes in birch number:MN117546), BpGRAS16 (GenBank number:MN117547) and BpGRAS19 (GenBank number:MN117548). ACT and TEF were used as the most stable reference genes in the control and 200 mM NaCl-treated samples, while GAPDH was used as the least stable reference gene. EF-1α, TUB, and their combinations were used as the most stable reference genes in different birch tissues, while GAPDH was used as the least stable reference gene. The same qRT-PCR reaction systems and conditions were used as stated above. Three parallel total RNA samples -ΔΔCt were applied to analyze the relative expression. Data processing as descried in 2 [29]. Results Specificity and amplification efficiency of qRT-PCR primers Agarose gel electrophoresis showed that single PCR products could be amplified from the 11 candidate reference genes (S1 Fig), indicating that the primers designed by Primer 5.0 software had strong specificity and fulfilled the requirements of qRT-PCR primers. The melting curves for all 11 candidate reference genes had a distinct single peak, and the three replications showed perfect repeatability (S2 Fig). The amplification efficiency of these reference genes ran- ged from 1.86 to 2.01 (Table 1), which met the requirements of the qRT-PCR experiment. Therefore, these subjects reference genes (ACT, TUA, TUB, TEF, 18S rRNA, EF1α, GAPDH, UBC, YLS8, SAND, and CDPK) could serve as the research subjects. Analysis of the expression level of the 11 candidate reference genes Ct value is one of the criteria used to measure gene expression level. A gene with a higher Ct value would have a lower gene expression level [29]. Ct values of the 11 candidate reference genes were unevenly distributed in the different treatments and tissues of birch (Fig 1). How- ever, the mean Ct value can illustrate the abundance of the transcripts from the 11 candidate reference genes (Table 1), which can be used to analyze the stability of reference genes in dif- ferent experimental samples. The mean Ct values of all the samples varied from 17.39 to 29.80. Among the 11 candidate reference genes in all samples, 18S rRNA had the highest expression level, with an average Ct ± SD of 17.39 ± 1.36. ACT, TUA, TUB, TEF, EF-1α, and YLS8 had relatively high expression level, with average Ct ± SD values of 22.49 ± 1.03, 22.80 ± 1.00, 23.23 ± 1.05, 22.28 ± 1.41, 23.09 ±1.96, and 23.47 ± 1.53, respectively. CDPK had the highest average Ct value, indicating that it had the lowest expression level (Fig 1). The lower the coefficient of variation (CV) value, the higher the stability of the candidate reference gene [30]. Among the 11 candidate reference genes in all samples, the CV value of TUA was the lowest, followed by TUB and ACT. GAPDH had a high CV value, which indicated that its expression was least stable (Table 1). Thus, the results indicated that the expression levels of 11 candidate reference genes were different among the different experimental conditions. GeNorm analysis The geNorm software can identify reference genes with better stabilities by calculating the average expression stability index (M value). The software will rank candidate reference genes expression stabilities according to their M values. The geNorm algorithm reveals the appropri- ate number of reference genes required for normalization to calculate the pairwise variation between the normalization factors NFn and NFn+1 of the two sequences (Vn/n+1). When Vn/n+1 is less than 0.15, n is the number of genes used in the normalization [31–32]. The lower the M value, the more stable the reference gene. Conversely, a higher M value indicates PLOS ONE | https://doi.org/10.1371/journal.pone.0225926 December 3, 2019 5 / 13 Selection of appropriate reference genes in birch Fig 1. Cycle threshold (Ct) value distribution of eleven candidate reference genes in different tissues and abiotic stress conditions in birch. Boxes contain the 5th and 95th quartiles of the Ct values of the different genes tested, median (central horizontal line) and minimal/maximal value (vertical bar). https://doi.org/10.1371/journal.pone.0225926.g001 worse stability. Genes with an M value less than 1.5 could be used as reference genes [33]. To obtain optimum reference genes for the different experimental conditions, total samples were divided into four groups including different tissues, normal, salt, and osmosis, as described in the section describing plant material and treatments. The results are shown in Fig 2. Ten of the eleven candidate reference genes in this study could be considered as appropriate reference genes with the M value less than 1.5, except for GAPDH in salt, different tissues, and total samples. Fig 2. Expression stability and ranking of eleven candidate reference genes as analyzed by geNorm. The black spots represent the average expression stability (M). The highest M value indicates the most unstable gene, while the lowest represents the most stable. The order from left to right indicates the stability ranking of the eleven candidate reference genes. The most variable is on the left while the most stable is on the right. https://doi.org/10.1371/journal.pone.0225926.g002 PLOS ONE | https://doi.org/10.1371/journal.pone.0225926 December 3, 2019 6 / 13 Selection of appropriate reference genes in birch In the different tissues of birch, TUB and EF-1α were the most stable reference genes. In normal conditions, ACT and TUA were the most stable. Under salt stress treatment, ACT and EF-1α were optimum reference genes. Under osmotic stress treatment, ACT and TUB were the most ideal reference genes. In total samples, ACT and TUA were the most ideal choices (Fig 2). In conclusion, the most appropriate reference genes were different in the different experimental conditions. However, ACT could be the best reference gene in all experimental conditions except in different tissues; GAPDH was the least stable reference gene in all the groups in this study. NormFinder analysis NormFinder software can select the most appropriate reference gene according to candidate reference genes expression steady values [34]. A reference gene with the least steady value (SV value) has the most stable expression. NormFinder can not only compare the expression differ- ences of candidate reference genes, but also compares the variation between samples. The results of gene stability ranking of candidate reference genes were generated and are shown in Table 2. TUB was the most stable reference gene in total group and in the normal conditions group. TEF was optimum reference gene under salt and osmotic stress. GAPDH was the least stable reference gene in all experimental groups. 18S rRNA was unstable in most groups, while it was optimum in the different tissues group. The above results are basically consistent with the geNorm analysis. Bestkeeper analysis Bestkeeper software is commonly used to choose appropriate reference genes; however, the number of candidate reference genes cannot exceed 10. Thus, we chose the 10 best candidate reference genes according to the results of geNorm and NormFinder. In BestKeeper, SD and R values are used to select ideal reference genes, however, the mean standard deviation (mSD) can be used to assess whether the expression of a gene is stable [35]. A gene with an SD value less than 1.0 can be considered as ideal candidate reference genes. At the same time, the closer the R value is to 1, the better the stability of the reference gene. The results of ranking the can- didate reference genes according to their SD values are shown in Table 3. In the different tissues group, it was obvious that TEF was the best reference gene, although its R value was far from satisfactory. In normal conditions, CDPK had the lowest SD value, Table 2. Expression stability values (SV value) of candidate reference genes under different treatments of birch, as calculated using Normfinder. Rank Different tissues Normal conditions Salt stress Osmotic stress Total Gene SV Gene SV Gene SV Gene SV Gene SV 1 18S rRNA 0.038 TUB 0.356 TEF 0.272 TEF 0.225 TUB 0.302 2 EF1α 0.083 EF1α 0.416 ACT 0.276 TUB 0.289 ACT 0.340 3 SAND 0.097 SAND 0.418 TUB 0.286 EF1α 0.387 TUA 0.341 4 ACT 0.099 ACT 0.503 EF1α 0.316 ACT 0.428 YLS8 0.562 5 TUB 0.211 YLS8 0.526 TUA 0.430 YLS8 0.526 UBC 0.741 6 YLS8 0.458 TUA 0.557 UBC 0.473 SAND 0.547 SAND 0.748 7 TEF 0.483 18S rRNA 0.694 SAND 0.501 TUA 0.624 TEF 0.806 8 TUA 0.532 CDPK 0.695 18S rRNA 0.777 CDPK 0.732 EF1α 0.948 9 UBC 1.117 TEF 0.745 YLS8 0.889 UBC 0.799 CDPK 1.091 10 CDPK 1.458 UBC 0.968 CDPK 0.907 18S rRNA 1.137 18S rRNA 1.097 11 GAPDH 2.226 GAPDH 1.685 GAPDH 2.328 GAPDH 1.438 GAPDH 1.399 https://doi.org/10.1371/journal.pone.0225926.t002 PLOS ONE | https://doi.org/10.1371/journal.pone.0225926 December 3, 2019 7 / 13 Selection of appropriate reference genes in birch Table 3. Expression stability values (standard deviation (SD) value and R value) of candidate reference genes under different treatments of birch, as determined by BestKeeper. Rank Different tissues Normal conditions Salt stress Osmotic stress Total Gene SD R value Gene SD R value Gene SD R value Gene SD R value Gene SD R value 1 TEF 0.070 0.704 CDPK 0.490 0.776 TEF 0.310 0.572 TUB 0.310 0.955 ACT 0.550 0.888 2 TUA 0.090 0.414 ACT 0.630 0.912 TUA 0.350 0.753 ACT 0.350 0.773 TUA 0.550 0.891 3 TUB 0.230 0.895 18S rRNA 0.650 0.694 ACT 0.420 0.952 TUA 0.440 0.697 TUB 0.550 0.881 4 EF1α 0.250 0.997 TUA 0.710 0.909 TUB 0.420 0.869 TEF 0.480 0.950 18S rRNA 0.690 0.356 5 ACT 0.380 0.903 TUB 0.740 0.950 EF1α 0.510 0.943 YLS8 0.570 0.755 TEF 0.710 0.662 6 18S rRNA 0.570 0.941 TEF 0.950 0.901 UBC 0.510 0.750 EF1α 0.630 0.944 YLS8 0.810 0.877 7 SAND 0.630 0.993 EF1α 0.970 0.957 SAND 0.630 0.873 SAND 0.640 0.856 SAND 0.880 0.793 8 UBC 0.710 0.149 YLS8 0.980 0.914 18S rRNA 0.730 0.348 CDPK 0.710 0.630 EF1α 0.930 0.776 9 YLS8 0.780 0.997 SAND 1.020 0.947 CDPK 0.760 0.879 UBC 1.030 0.632 UBC 0.970 0.822 10 CDPK 0.780 -0.174 UBC 1.120 0.848 YLS8 0.860 0.853 18S rRNA 1.230 0.765 CDPK 1.090 0.711 https://doi.org/10.1371/journal.pone.0225926.t003 followed by ACT, and the R value of ACT was closer to 1 than that of CDPK. On the whole, CDPK was more appropriate to serve as a candidate reference gene under the normal experi- mental conditions used. Under salt stress conditions, SD values of all the candidate reference genes were less than 1.0. Thus, the 10 tested candidate reference genes could be used as refer- ence genes, however, the R values of ACT and EF1α are closer to 1, so we consider choosing them. TEF was the best candidate reference gene, which agreed with the results of NormFin- der. ACT and EF1α had R values that closer to 1. Under osmotic stress conditions, TUB was the best choice because of its ideal SD and R values. Ultimately, ACT, TUA, and TUB had simi- lar stabilities according to the results in all samples. These three candidate reference genes were suitable as candidate reference genes, which agreed with the results of geNorm. Validation of reference genes To verify the selected reference genes, the most and least stable reference genes were used to assess the relative expression level of target BpGRAS genes after normalization under salt stress and different tissues. The GRAS transcription factor family is one of the largest families of transcription factors and are involved in regulating the expression of several target function genes in plants biotic and abiotic responses, including that to salt. High-salt and drought stress could induce the expression of SCL7 in Populus euphratica, which belongs to the GRAS tran- scription factor family [36]. Eight GRAS transcription factor genes were upregulated in differ- ent tissue of Dendrobium catenatum following exposure to heat and salt stresses [37]. Our previous research showed that transiently overexpressed BpGRASs had significantly increased salt tolerance in birch (unpublished). We used the most stable and least stable candidate reference genes according to our results to normalize the expression level of BpGRAS genes. The expression patterns of the BpGRAS genes were markedly different when the most and least stable reference genes were used to nor- malize their expression in salt stress and different tissues (Fig 3). When the ideal reference genes ACT and TEF were used for normalization under treatment with 200 mM NaCl, the expression level of the BpGRAS genes were significantly enhanced compared with that in the control group at most time points, and their expression patterns were similar. However, we could not draw the same conclusion when the least stable reference gene was used (Figs 3 and 4). When the ideal reference genes ACT and TEF were used as inter- nal controls for samples treated with 200 mM NaCl, BpGRAS1 had at the highest transcript level at 48 h after salt treatment. However, when the least stable reference gene GAPDH was PLOS ONE | https://doi.org/10.1371/journal.pone.0225926 December 3, 2019 8 / 13 Selection of appropriate reference genes in birch Fig 3. Relative quantification of targeted BpGRAS gene expression levels in birch using validated reference genes. Samples of leaves collected at 0, 3, 6, 12, 24, and 48 h under salt stress, respectively. A1-C1: Samples treated with 200 mM NaCl using ACT as internal the reference gene to normalize the expression of BpGRAS genes. A2-C2: Samples treated with 200 mM NaCl using TEF as the internal reference gene to normalize the expression of BpGRAS genes. A3-C3: Samples treated with 200 mM NaCl using GAPDH as the internal reference gene to normalize the expression of BpGRAS gene. represented 0.01 < P < 0.05. https://doi.org/10.1371/journal.pone.0225926.g003 used to normalize the expression level of BpGRAS1, BpGRAS1 showed it was hardly expressed at 48 h. The results are the same as follows, when the ideal reference genes ACT and TEF were used for normalization after treatment of 200 mM NaCl, BpGRAS16 and BpGRAS19 showed higher transcript levels at 48 h after salt treatment; however, it was hardly expressed at 48 h when the least stable reference gene, GAPDH, was used for normalization. Simultaneously, T- test showed that the expression level of NaCl treatment group was significantly higher than that of the control. The most stable reference genes for use in different tissues were EF1α and TUB and the least stable reference gene was GAPDH. These three genes were used to normalize the expres- sions of the BpGRAS genes. The combination of EF1α and TUB was also used for normaliza- tion. The results are shown in Fig 4. When the appropriate reference genes or their combination were used, the expression of BpGRAS1 in root was the highest, followed by that in the stem, with the lowest expression in the leaf. The expression of BpGRAS16 was highest in the stem, followed by leaf and then the root. The expression of BpGRAS19 in the stem and root were the essentially the same. However, when the least stable reference gene was used for nor- malization, the BpGRAS genes were hardly expressed in the stem and root. Obviously, the expression patterns of the BpGRAS genes were similar after using the appropriate reference genes and their combinations for normalization. However, the expression patterns of BpGRAS genes showed large differences when the least stable reference gene was used. Fig 4. Expression levels in birch using the validated reference genes. Samples of leaves, stems, and roots in normal conditions assessed using qRT-PCR with different internal reference gene to normalize the expression of BpGRAS genes. represented 0.01 < P < 0.05. https://doi.org/10.1371/journal.pone.0225926.g004 PLOS ONE | https://doi.org/10.1371/journal.pone.0225926 December 3, 2019 9 / 13 Selection of appropriate reference genes in birch In summary, using different reference genes for normalization could produce different experimental results. Therefore, selecting the ideal reference genes according to different experimental conditions and different tissues is very importance. Discussion Determining a gene’s expression level is an important step to understand the function of the gene product in biological processes during environmental stress, and in developmental and cellular processes [38]. To ensure the accuracy of relative RT-PCR, specific PCR conditions and an appropriate internal controls must be determined. Thus, the selection of internal refer- ence genes that show stable expression in all tissues and under the experimental conditions being investigated plays a key role in determining stress-related gene expression under differ- ent abiotic stress conditions. Identifying suitable internal reference genes in birch will encour- age gene expression studies in this tree species. There are many reports concerning reference gene selection in plant, including in Salicor- nia europaea [12], cucumber [15]. However, until now, there have been no systematic reports about reference genes selection in birch. Therefore, the present study was designed to identify and validate suitable reference genes for gene expression analysis in birch. In this study, 11 candidate reference genes that have been frequently used in previous stud- ies were evaluated, and three data analysis tools (geNorm, NormFinder, BestKeeper), which are specialized for reference gene selection, were used. The expression levels of the 11 candi- date reference genes were significantly different among the different samples (Fig 1). The three analysis tools have different algorithms; therefore, the ranking of these 11 candidate reference genes according to their expression stability varied. The differences in the statistical theories used by the three pieces of software would explain the different results. Using correlation anal- ysis (R ) 11 candidate genes between M values vs SV values (ranged from 0.0022 to 0.8169), SV values vs SD values (ranged from 0.0048 to 0.8232), M values vs SD values (ranged from 0.0068 to 0.9159), which indicated the data could not establish the expression stability. We should select the reference genes with better expression stability as identified by all three tools. Therefore, the rankings of candidate reference genes calculated by these three algorithms are listed (S1 Table). In general, EF1α was the best reference gene in different tissues. ACT and TEF are the best for salt-stressed sample and TUB alone is the best for osmotic-stressed sample. Meanwhile, ACT was also the most appropriate reference gene in normal conditions and in all the samples. Previously GAPDH was identified as the optimum reference gene in salt-treated leaves, Cd-treated roots, cold-treated leaves and roots, and PEG-treated leaf samples in Carex rigescens [20]. However, GAPDH was the least stable reference gene in all the experimental conditions in the present study, despite being one of the most commonly used reference genes in previous studies. Suitable reference genes have been confirmed in many plant species in different tissues and under abiotic stress. In Salicornia europaea, ACT (Actin) and GAPDH were the optimum com- bination of internal reference genes to study gene expression under drought stress [12]. In rice, UBQ5 and eEF1α was most stable as reference genes in different tissues [39]. In tomato, the expression stabilities of GAPDH and phosphoglycerate kinase (PGK) were ranked as the top during light stress but were poorly ranked during N and cold stress [40]. In cucumber, TUA was considered as an appropriate reference gene in different tissues. EF1α was identified as a suitable reference gene for abiotic stress treatment [15]. Thus, appropriate reference genes differ among different plants and according to the experimental conditions. In conclusion, dif- ferent reference genes should be selected according to different experimental conditions to obtain accurate results. PLOS ONE | https://doi.org/10.1371/journal.pone.0225926 December 3, 2019 10 / 13 Selection of appropriate reference genes in birch Conclusion In this study, ACT was identified as the optimum reference gene in all experimental groups, except in the different tissues group. GAPDH was the least stable candidate reference gene in all experimental conditions. This study provided appropriate reference genes for expression studies in birch, which will be beneficial for more accurate relative quantification of mRNA expression in birch for different tissues, in normal conditions, and under salt and osmotic stress conditions. Supporting information S1 Fig. The results of 1.5% agarose gel electrophoresis after 30 cycles PCR. (TIF) S2 Fig. Dissociation curves of the qRT-PCR amplicons. (TIF) S1 Table. Comprehensive rankings of the stability of the reference genes by three algo- rithms. GN: Ranking of candidate reference genes calculated by geNorm. NF: Ranking of can- didate reference genes calculated by NormFinder. BK: Ranking of candidate reference genes calculated by Bestkeeper (DOCX) Author Contributions Conceptualization: Ziyi Li, Yucheng Wang, Xiaoyu Ji. Data curation: Ziyi Li. Formal analysis: Ziyi Li, Huijun Lu, Zihang He. Funding acquisition: Xiaoyu Ji. Investigation: Huijun Lu. Methodology: Xiaoyu Ji. Project administration: Chao Wang, Yucheng Wang. Software: Ziyi Li, Huijun Lu, Zihang He. Supervision: Yucheng Wang. Validation: Zihang He, Chao Wang. Visualization: Huijun Lu, Zihang He, Chao Wang. Writing – original draft: Ziyi Li, Huijun Lu. Writing – review & editing: Ziyi Li. References 1. Morandi A, Zhaxybayeva O, Gogarten JP, Graf J. Evolutionary and diagnostic implications of intrage- nomic heterogeneity in the 16S rRNA gene in Aeromonas strains. J Bacteriol. 2005; 187(18):6561– 6564. https://doi.org/10.1128/JB.187.18.6561-6564.2005 PMID: 16159790 2. Garrido-Maestu A, Chapela MJ, Vieites JM, Cabado AG. Lolb gene, a valid alternative for qPCR detec- tion of Vibrio cholerae in food and environmental samples. Food Microbiology. 2015; 46(46):535–540. PLOS ONE | https://doi.org/10.1371/journal.pone.0225926 December 3, 2019 11 / 13 Selection of appropriate reference genes in birch 3. Godhe A, Otta SK, Rehnstam-Holm AS, Karunasagar I. Polymerase chain reaction in detection of Gym- nodinium mikimotoi and Alexandrium minutum in field samples from southwest India. Mar Biotechnol (NY). 2001; 3(2):152–162. https://doi.org/10.1007/s101260000052 PMID: 14961378 4. Steeples LR, Guiver M, Jones NP. Real-time PCR using the 529 bp repeat element for the diagnosis of atypical ocular toxoplasmosis. Br J Ophthalmo. 2016; 100(2): 200–203. 5. Zhu J, Zhang L, Li W, Han S, Yang W, Qi L. Reference gene selection for quantitative real-time PCR normalization in Caragana intermedia under different abiotic stress conditions. PLoS ONE. 2013; 8: e53196. https://doi.org/10.1371/journal.pone.0053196 PMID: 23301042 6. Dean JD, Goodwin PH, Hsiang T. Comparison of relative RT-PCR and northern blot analyses to mea- sure expression ofβ-1,3-glucanase in nicotiana benthamiana, infected with colltotrichum destructivum. Plant Molecular Biology Reporter. 2002; 20(4):347–356. 7. Gare EM, Divjak M, Bailey MJ, Walters EH. Beta-Actin and GAPDH housekeeping gene expression in asthmatic airways is variable and not suitable for normalising mRNA levels. Thorax. 2002; 57(9):765– 770. https://doi.org/10.1136/thorax.57.9.765 PMID: 12200519 8. Migocka M, Papierniak A. Identification of suitable reference genes for studying gene expression in cucumber plants subjected to abiotic stress and growth regulators. Mol. Breed. 2011; 28:343–357. 9. Reddy PS, Reddy DS, Sharma KK, Bhatnagar-Mathur P, Vadez V. Cloning and validation of reference genes for normalization of gene expression studies in pearl millet by quantitative real-time PCR. Plant Gene. 2015; 1:35–42. 10. Lv S, Jiang P, Chen X, Fan P, Wang X, Li Y. Multiple compartmentalization of sodium conferred salt tol- erance in Salicornia europaea. Plant Physiol Biochem. 2011; 51:47–52. https://doi.org/10.1016/j. plaphy.2011.10.015 PMID: 22153239 11. Ma J, Zhang M, Xiao X, You J, Wang J, Wang T, et al. Global transcriptome profiling of Salicornia euro- paea L. shoots under NaCl treatment. PLoS ONE. 2013; 8:e65877. https://doi.org/10.1371/journal. pone.0065877 PMID: 23825526 12. Xiao X, Ma J, Wang J, Wu X, Li P, Yao Y. Validation of suitable reference genes for gene expression analysis in the halophyte Salicornia europaea by real-time quantitative PCR. Front Plant Sci.2014; 5:788. https://doi.org/10.3389/fpls.2014.00788 PMID: 25653658 13. Kiarash JG, Dayton WH, Amirmahani F, Mehdi MM, Zaboli M, Nazari M.Selection and validation of ref- erence genes for normalization of qRT-PCR gene expression in wheat (Triticum durum L.) under drought and salt stresses. J Genet. 2018; 97(5): 1433–1444. PMID: 30555091 14. Tang X, Wang H, Shao C, Shao H. Reference Gene Selection for qPCR Normalization of Kosteletzkya virginica under Salt Stress. Biomed Res Int. 2015: 823806. https://doi.org/10.1155/2015/823806 PMID: 15. Wan H, Zhao Z, Qian C, Sui Y, Malik AA, Chen J. Selection of appropriate reference genes for gene expression studies by quantitative real-time polymerase chain reaction in cucumber. Anal Biochem. 2010; 399(2):257–261. https://doi.org/10.1016/j.ab.2009.12.008 PMID: 20005862 16. Zhu X, Li X, Chen W, Chen J, Lu W, Chen L, et al. Evaluation of new reference genes in papaya for accurate transcript normalization under different experimental conditions. PLoS One. 2012; 7(8): e44405. https://doi.org/10.1371/journal.pone.0044405 PMID: 22952972 17. Czechowski T, Stitt M, Altmann T, Udvardi MK, Scheible WR. Genome-wide identification and testing of superior reference genes for transcript normalization in Arabidopsis. Plant Physiol.2005; 139(1):5–17. https://doi.org/10.1104/pp.105.063743 PMID: 16166256 18. Wan Q, Chen S, Shan Z, Yang Z, Chen L, Zhang C, et al. Stability evaluation of reference genes for gene expression analysis by RT-qPCR in soybean under different conditions. PLoS One.2017; 12(12): e0189405. https://doi.org/10.1371/journal.pone.0189405 PMID: 29236756 19. Shukla P, Reddy RA, Ponnuvel KM, Rohela GK, Shabnam AA, Ghosh MK, et al. Selection of suitable reference genes for quantitative real-time PCR gene expression analysis in Mulberry (Morus alba L.) under different abiotic stresses. Mol Biol Rep. 2019; 6(2):1809–1817. 20. Zhang K, Li M, Cao S, Sun Y, Long R, Kang J, et al. Selection and alidation of reference genes for target gene analysis with quantitative real-time PCR in the leaves and roots of Carex rigescens under abiotic stress. Ecotoxicol Environ Saf. 2019; 168:127–137. https://doi.org/10.1016/j.ecoenv.2018.10.049 PMID: 30384160 21. Kang D, Guo Y, Ren C, Zhao F, Feng Y, Han X, et al. Population structure and spatial pattern of main tree species in secondary Betula platyphylla forest in Ziwuling Mountains, China. Sci Rep. 2014; 4:6873. https://doi.org/10.1038/srep06873 PMID: 25362993 22. Miao L, Qin X, Gao L, Li Q, Li S, He C, et al. Selection of reference genes for quantitative real-time PCR analysis in cucumber (Cucumis sativus L.), pumpkin (Cucurbita moschata Duch.) and cucumber–pump- kin grafted plants. Peer J. 2019; 7:e6536. https://doi.org/10.7717/peerj.6536 PMID: 31024757 PLOS ONE | https://doi.org/10.1371/journal.pone.0225926 December 3, 2019 12 / 13 Selection of appropriate reference genes in birch 23. Ma R, Xu S, Zhao Y, Xia B, Wang R. Selection and Validation of Appropriate Reference Genes for Quantitative Real-Time PCR Analysis of Gene Expression in Lycoris aurea. Front Plant Sci. 2016; 7:536. https://doi.org/10.3389/fpls.2016.00536 PMID: 27200013 24. Pfaffl MW. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res. 2001; 29:e45. https://doi.org/10.1093/nar/29.9.e45 PMID: 11328886 25. Vandesompele J, De PK, Pattyn F, Poppe B, Van RN, De PA, et al. Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol. 2002; 3(7):RESEARCH0034. 26. Andersen CL, Jensen JL, Orntoft TF. 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.2004; 64:5245–5250. https://doi.org/10.1158/0008- 5472.CAN-04-0496 PMID: 15289330 27. Ramakers C, Ruijter JM, Deprez RH, Moorman AF. Assumption-free analysis of quantitative real-time polymerase chain reaction (PCR) data. Neurosci Lett. 2003; 339(1):62–66. https://doi.org/10.1016/ s0304-3940(02)01423-4 PMID: 12618301 28. Pfaffl MW, Tichopad A, Prgomet C, Neuvians TP. Determination of stable housekeeping genes, differ- entially regulated target genes and sample integrity: BestKeeper-Excel-Based tool using pair-wise cor- relations. Biotechnol Lett.2004; 26:509–515. https://doi.org/10.1023/b:bile.0000019559.84305.47 PMID: 15127793 29. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2-ΔΔCt method. Methods.2001; 25:402–408. https://doi.org/10.1006/meth.2001.1262 PMID: 30. Ni X, Qi J, Zhang G, Xu J, Tao A, Fang P, et al. Selection of reliable reference genes for quantitative real-time PCR gene expression analysis in Jute (Corchorus capsularis) under stress treatments. Front Plant Sci.2015; 6:848. https://doi.org/10.3389/fpls.2015.00848 PMID: 26528312 31. Wang JJ, Han S, Yin W, Xia X, Liu C. Comparison of Reliable Reference Genes Following Different Hor- mone Treatments by Various Algorithms for qRT-PCR Analysis of Metasequoia. International Journal of Molecular Sciences. 2018. 20(1):34. 32. Wei Y, Liu Q, Dong H, Zhou Z, Hao Y, Chen X, et al. Selection of Reference Genes for Real-Time Quan- titative PCR in Pinus massoniana.Post Nematode Inoculation. PLoS One. 2016; 11(1):e0147224. https://doi.org/10.1371/journal.pone.0147224 PMID: 26800152 33. Vandesompele J, De PK, Pattyn F, Poppe B, Van RN, De PA, et al. Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol. 2002; 3(7): RESEARCH0034. 34. Andersen CL, Jensen JL,Orntoft TF. 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.2004; 64:5245–5250. https://doi.org/10.1158/0008- 5472.CAN-04-0496 PMID: 15289330 35. Zeng X, Ling H, Chen X, Gu S. Genome-wide identification, phylogeny and function analysis of GRAS gene family in Dendrobium catenatum (Orchidaceae). Gene. 2019; 705:5–15. https://doi.org/10.1016/j. gene.2019.04.038 PMID: 30999026 36. Ma HS, Xia XL, Yin WL. Cloning and analysis of SCL7 gene from Populus euphratica. Beijing Linye Daxue Xuebao (Journal of Beijing Forestry University).2011; 33:1–10. 37. Wang Y, Liu Z, Wu Z, Li H, Wang W, Cui X, et al. Genome-wide identification and expression analysis of GRAS family transcription factors in tea plant (Camellia sinensis). Sci Rep. 2018; 8(1):3949. https:// doi.org/10.1038/s41598-018-22275-z PMID: 29500448 38. Hu R, Fan C, Li H, Zhang Q, Fu YF. Evaluation of putative reference genes for gene expression normal- ization in soybean by quantitative real-time RT-PCR. BMC Molecular Biology.2009; 10:93. https://doi. org/10.1186/1471-2199-10-93 PMID: 19785741 39. Jain M, Nijhawan A, Tyagi AK, Khurana JP. Validation of housekeeping genes as internal control for studying gene expression in rice by quantitative real-time PCR. Biochem Biophys Res Commun. 2006; 345(2):646–651. https://doi.org/10.1016/j.bbrc.2006.04.140 PMID: 16690022 40. Løvdal T, Lillo C. Reference gene selection for quantitative real-time PCR normalization in tomato sub- jected to nitrogen, cold, and light stress. Analytical Biochemistry.2009; 387(2):238–242. https://doi.org/ 10.1016/j.ab.2009.01.024 PMID: 19454243 PLOS ONE | https://doi.org/10.1371/journal.pone.0225926 December 3, 2019 13 / 13 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png PLoS ONE Public Library of Science (PLoS) Journal

Selection of appropriate reference genes for quantitative real-time reverse transcription PCR in Betula platyphylla under salt and osmotic stress conditions

PLoS ONE , Volume 14 (12) – Dec 3, 2019
Download PDF
13 pages

Loading next page...
 
/lp/public-library-of-science-plos-journal/selection-of-appropriate-reference-genes-for-quantitative-real-time-3LU8A17zNx

References (78)

Publisher
Public Library of Science (PLoS) Journal
Copyright
Copyright: © 2019 Li et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability: All relevant data are within the manuscript and its Supporting Information files. Funding: This work was supported by the National Natural Science Foundation of China (No. 31500535), and The Heilongjiang Province Science Foundation (No. QC2018017). Competing interests: The authors have declared that no competing interests exist.
eISSN
1932-6203
DOI
10.1371/journal.pone.0225926
Publisher site
See Article on Publisher Site

Abstract

Selecting appropriate reference genes is vital to normalize gene expression analysis in birch (Betula platyphylla) under different abiotic stress conditions using quantitative real- OPENACCESS time reverse transcription PCR (qRT-PCR). In this study, 11 candidate birch reference genes (ACT, TUA, TUB, TEF, 18S rRNA, EF1α, GAPDH, UBC, YLS8, SAND, and CDPK) Citation: Li Z, Lu H, He Z, Wang C, Wang Y, Ji X (2019) Selection of appropriate reference genes for were selected to evaluate the stability of their expression in different tissues and under dif- quantitative real-time reverse transcription PCR in ferent abiotic stress conditions. Three statistical algorithms (GeNorm, NormFinder, and Betula platyphylla under salt and osmotic stress BestKeeper) were used to analyze the stability of the 11 candidate reference genes to iden- conditions. PLoS ONE 14(12): e0225926. https:// tify the most appropriate one. The results indicated that EF-1α was the most stable refer- doi.org/10.1371/journal.pone.0225926 ence gene in different birch tissues, ACT was the most stable reference gene for normal Editor: Mayank Gururani, United Arab Emirates conditions, ACT and TEF were the most stable reference genes for salt stress treatment, University, UNITED ARAB EMIRATES TUB was the most stable reference gene for osmotic stress treatment, and ACT was the Received: September 9, 2019 most appropriate choice in all samples of birch. In conclusion, the most appropriate refer- Accepted: November 15, 2019 ence genes varied among different experimental conditions. However, in this study, ACT Published: December 3, 2019 was the optimum reference gene in all experimental groups, except in the different tissues Peer Review History: PLOS recognizes the group. GAPDH was the least stable candidate reference gene in all experimental conditions. benefits of transparency in the peer review In addition, three stress-induced genes (BpGRAS1, BpGRAS16, and BpGRAS19) were process; therefore, we enable the publication of chosen to verify the stability of the selected reference genes in different tissues and under all of the content of peer review and author salt stress. This study laid the foundation for the selection of appropriate reference gene(s) responses alongside final, published articles. The editorial history of this article is available here: for future gene expression pattern studies in birch. https://doi.org/10.1371/journal.pone.0225926 Copyright:© 2019 Li et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and Introduction source are credited. Quantitative real-time reverse transcription polymerase chain reaction (qRT-PCR) has Data Availability Statement: All relevant data are become a common technique to detect gene expression levels in the field of molecular biology within the manuscript and its Supporting Information files. because of its speed, high efficiency, labor-saving, and sensitivity. It has been widely used in PLOS ONE | https://doi.org/10.1371/journal.pone.0225926 December 3, 2019 1 / 13 Selection of appropriate reference genes in birch Funding: This work was supported by the National basic biological research [1], food spoilage detection [2], environmental monitoring [3], and Natural Science Foundation of China (No. diagnosis and treatment of disease [4]. QRT-PCR has strict requirements on the purity and 31500535), and The Heilongjiang Province Science quality of RNA, reverse transcription efficiency, the specificity of the primers, PCR amplifica- Foundation (No. QC2018017). tion efficiency, and the selection of internal reference genes [5]. Reference genes can have a Competing interests: The authors have declared large influence on the results of relative expression level determination. Thus, selecting ideal that no competing interests exist. reference genes is vital to obtain reliable results. One or more internal reference genes are typi- cally required to normalize qRT-PCR data to obtain more accurate results. Over the course of an experiment, a reliable internal control should show a minimal change in expression, whereas the expression of a gene of interest may change greatly. Housekeeping genes are often selected because they encode proteins essential to cell viability and therefore are assumed to be expressed similarly in different cell types [6]. Commonly used internal con- trols include actin (ACT), beta-tubulin (TUB), elongation factor 1 alpha (EF1-α), and glyceral- dehyde-3-phosphate dehydrogenase (GAPDH) [7–8]. However, increasing numbers of reports have confirmed that the expression of housekeeping genes can vary in different tissues and may be affected by the experimental conditions. For instance, TUB is a wildly used reference gene; a study showed that it was unsuitable as a reference gene in different tissues of Pennise- tum glaucum [9]. The TUA (α-tubulin) gene was selected as reference gene in Salicornia euro- paea under salt stress [10–11]; whereas it was not considered as an ideal reference gene in S. europaea under drought treatment [12]. ACT was selected as a reference gene in wheat and Kosteletzkya virginica under salt stress, but was not suitable as a reference gene in papaya under many experimental conditions or in Salicornia europaea and cucumber under salt stress [12–16]. Therefore, selecting the appropriate reference genes according to different tissues and different experimental conditions is essential. Studies on the evaluation and validation of reference genes have been conducted in many plant species, especially in herbaceous plants, such as Arabidopsis thaliana [17], Glycine max [18]. Similar studies have been carried out in certain woody plants, such as Morus alba L. [19], Carex rigescens [20]. However, the systematic evaluation and validation of appropriate refer- ence genes in birch (Betula platyphylla) under different abiotic stress conditions has not been reported. Betula platyphylla, which is a species of deciduous hardwood, is widely distributed in the mid-high mountains of warm temperate regions in the world, including northern China, Rus- sian Far East, Siberia, Mongolia, Northern Korea, and Japan [21]. With the continuous expan- sion of drought and saline areas worldwide, it is of great significance to cultivate new salt tolerant and drought resistant varieties of birch. Therefore, selection of internal reference genes in birch with stable expression under different abiotic stress conditions plays a key role in studies on stress-related gene expression. The analysis results of qRT-PCR were affected to varying degrees of influence in reference genes choosing. Some studies show how to choose genes as reference genes to normalized expression data. [22]. Therefore, in this study, eleven candidate reference genes (ACT, TUA, TUB, TEF (Translation elongation factor)), 18S rRNA (18S ribosomal RNA), EF1α, GAPDH, UBC (Ubiquitin-conjugating enzyme E2), YLS8 (Thioredoxin-like protein YLS8), SAND (S- adenosyl methionine decarboxylase), and CDPK (CDPK-related kinase) were selected to eval- uate their expression stability in different tissues and under abiotic stress conditions in birch. Three different statistical tools (GeNorm, NormFinder, and BestKeeper) were used to analyze the stability of 11 candidate reference genes and identify the most appropriate reference gene. Our results will facilitate of appropriate reference genes selection for the development of gene expression pattern studies in birch. PLOS ONE | https://doi.org/10.1371/journal.pone.0225926 December 3, 2019 2 / 13 Selection of appropriate reference genes in birch Materials and methods Plant material and stress treatments In this experiment, open pollinated northeast white birch (B. platyphylla) seeds from the North- east Forestry University were planted in plastic pots with mixture of turf peat and sand (v/v 3:1), -2 -1 in a growth house where the conditions kept at 70–75% relative humidity, 400 μmol�m s light intensity, a 16 h light/8 h dark photocycle at 25±2˚C. The 8-week-old birch plants grown in soil were watered with a solution of 2 L 200 mM NaCl or 2 L 300 mM mannitol (distilled water served as the control) for 3, 6, 12, 24, and 48 h, respectively. The root, stem, and leaf tissues from 6 birch seedlings were collected after each treatment. Three independent replications were performed. All the samples were placed at −80˚C after freezing in liquid nitrogen for further study. Total RNA isolation and cDNA synthesis The total RNA of all samples was isolated using a Universal Plant Total RNA Extraction Kit (Bioteke, Beijing, China) following the manufacturer’s instructions. RNA integrity was checked using 1.5% agarose gel electrophoresis, and the RNA concentration and purity were determined using a NanoDrop 2000 Spectrophotometer (NanoDrop, Thermo Scientific, Wal- tham, MA, USA). All the RNA samples with A260/A280 ratios = 1.9–2.1 and A260/A230 ratios > 2.0 were used for further study. Then RNA (1 μg) was reversely transcribed into cDNA using a TransScript One-Step gDNA Removal and cDNA Synthesis SuperMIX (Trans- Gen Biotech, Beijing, China) following the manufacturer’s instructions. The synthesized cDNA was stored at −20˚C for further study. Candidate reference gene selection and primer design Eleven candidate reference genes that are commonly used in qRT-PCR were chosen for this study ACT, TUA, TUB, TEF, 18S rRNA, EF1α, GAPDH, UBC, YLS8, SAND, and CDPK [22– 23]. The GenBank accession numbers for the 11 genes, which were obtained from the birch genome data (unpublished), are listed in Table 1. Specific primers for the reference genes were designed using Primer premier 5.0 (Premier Biosoft International, Palo Alto, CA, USA). The primer sequences for qRT-PCR are listed in Table 1. Quantitative real time PCR A TransStart Top Green qPCR SuperMix kit (TransGen Biotech, Beijing, China) was used for qRT-PCR. All qRT-PCR reactions were carried out on Qtower G (Analytik, Jena, Germany). The qRT-PCR reaction system included: 10 μL of 2 × TransStart Top Green Realtime qPCR SuperMix, 0.4 μL upstream primers, and 0.4 μL downstream primers (10 μM), 2 μL of cDNA templates (100 ng), and 7.2 μL of RNAase-free ddH O with a volume of 20 μL. The qRT-PCR reaction conditions were as follows: 94˚C for 30 s; 45 cycles at 94˚C for 12 s, 60˚C for 30 s, and 72˚C for 40 s; and 79˚C for 1 s for plate reading. The melting curve was constructed to verify the specificity of each sequence-specific primer. The PCR amplification efficiency (E) was calculated as E = −1 + 10 (−1/ slope) [24], where the slope was derived from a standard curve generated by 10-fold serial dilu- tions of the mixed cDNA for each primer pair. Three independent replications were performed. Data analysis SigmaPlot 10 software (Systat Software, Inc., San Jose, CA, USA) was used to show the cycle threshold (Ct) value distribution of all samples, which can reflect clearly the average expression level of the genes in the samples. Threshold for the Ct values is the machine setting. The aver- age Ct value was calculated using three biological replications. GeNorm [25] and NormFinder PLOS ONE | https://doi.org/10.1371/journal.pone.0225926 December 3, 2019 3 / 13 Selection of appropriate reference genes in birch Table 1. Primer sequences and amplification parameters of 11 candidate reference genes from birch. Gene symbol Gene name Accession Primers Tm E Amplicon Mean Ct SD CV (%) 0 0 no. (5 -3 ) length (bp) 18S rRNA 18S ribosomal RNA MK388236 F: AACGAACGAGACCTCAGCCT 61.18 1.97 171 17.39 1.36 14.01 R: ACTCGTTGAATACATCAGTG 53.09 GAPDH Glyceraldehyde-3-phosphate dehydrogenase MK388226 F: AAGCTCAATGGCATTGCACT 58.74 1.89 205 26.87 2.33 15.64 R: TGGAAGAAACATCAGTGCAC 56.26 ACT Actin MK388227 F: TGAGAAGAGCTATGAGTTGC 54.89 1.98 204 22.49 1.03 8.26 R: GTAGATCCACCACTAAGCAC 55.27 TUA Alpha-tubulin MK388228 F: ATATCATCCTTGACAACTTC 50.18 1.94 210 22.80 1.00 7.92 R: GATGCCACATTTGAAGCCAG 57.71 TUB beta-tubulin MK388229 F: GTTAGCGAGCAGTTTACAGC 57.21 2.01 195 23.22 1.05 8.10 R: ACCAACAACCTCTTCTTCTT 54.39 UBC Ubiquitin-conjugating enzyme E2 MK388230 F: CAAATGACAGTCCCTATGCT 55.12 1.99 219 26.05 1.81 12.48 R: GGTCAGTAAGCAATGAACAT 53.58 EF-1α Elongation factor 1-alpha MK388231 F: GACAACGTTGGCTTCAACGT 59.62 1.92 203 23.09 1.96 15.27 R: GCAAACTTCACAGCAATGTG 56.43 TEF Translation elongation factor MK388232 F: GTACAATGATGAGAACACTG 51.64 1.96 201 22.28 1.41 11.38 R: TCCATGTCATAGCACTCATC 54.64 SAND Sadenosyl methionine decarboxylase MK388233 F: GCATCTAGGACAACCTAGAG 54.38 1.86 213 26.19 1.66 11.40 R: CCACTATCATGCATAGAAGC 53.65 YLS8 Thioredoxin-like protein YLS8 MK388234 F: GCATCTGTTGCTGAGACAAT 56.42 1.90 216 23.47 1.53 11.71 R: CTCCACAATATCAATGAACT 50.48 CDPK CDPK-related kinase MK388235 F: GAAGATGAGCTCATCTACCT 53.66 1.93 212 29.80 2.06 12.45 R: TAAGTACTGATTGCAGCAGC 55.85 Tm, Melting Temperature; E, PCR amplification efficiency; Ct, cycle threshold; CV, coefficient of variation; SD, standard deviation https://doi.org/10.1371/journal.pone.0225926.t001 [26] software were used to analyze the stabilities of the candidate reference genes’ expression. The mean Ct value must be transformed into relative expression level based on the formula -ΔCt -ΔCt E (-ΔCt = Ct value of each sample-the minimum Ct value) [27]. The E value can be ana- lyzed by GeNorm and NormFinder to obtain the candidate reference gene’s expression stabil- ity. GeNorm can determine the stability of each candidate reference gene through calculating the M value according to its expression. When the M value is less than 1.5, the gene can be used as an appropriate reference gene. Moreover, the lower the M value, the more stable the expression of the candidate reference gene. NormFinder can rank the stability of candidate reference genes expression. Moreover, it can identify the best reference gene among all candi- date reference genes. The BestKeeper [28] program analyzed the pairing correlation under a given set of experimental conditions using the Ct values of the reference genes in each group. According to the standard deviation (SD) and coefficient of correlation of the candidate genes R value, the expression stability of each reference gene was evaluated in the BestKeeper pro- gram. A reference gene with a lower SD value and an R value closer to 1 would have stable expression. Correlation analysis (R ) using SPSS (19.0) between M values vs SV values, SV value vs SD values, M values vs SD values for determine its data robustness. A simple T test using SPSS to illustrate the significance of data differences. Differences were considered to be significant if P < 0.05. represented 0.01 < P < 0.05. Validation of reference genes The identified optimum reference genes and the least stable reference genes in salt or in differ- ent tissues were used to normalize the relative expression levels of BpGRAS1 (GenBank PLOS ONE | https://doi.org/10.1371/journal.pone.0225926 December 3, 2019 4 / 13 Selection of appropriate reference genes in birch number:MN117546), BpGRAS16 (GenBank number:MN117547) and BpGRAS19 (GenBank number:MN117548). ACT and TEF were used as the most stable reference genes in the control and 200 mM NaCl-treated samples, while GAPDH was used as the least stable reference gene. EF-1α, TUB, and their combinations were used as the most stable reference genes in different birch tissues, while GAPDH was used as the least stable reference gene. The same qRT-PCR reaction systems and conditions were used as stated above. Three parallel total RNA samples -ΔΔCt were applied to analyze the relative expression. Data processing as descried in 2 [29]. Results Specificity and amplification efficiency of qRT-PCR primers Agarose gel electrophoresis showed that single PCR products could be amplified from the 11 candidate reference genes (S1 Fig), indicating that the primers designed by Primer 5.0 software had strong specificity and fulfilled the requirements of qRT-PCR primers. The melting curves for all 11 candidate reference genes had a distinct single peak, and the three replications showed perfect repeatability (S2 Fig). The amplification efficiency of these reference genes ran- ged from 1.86 to 2.01 (Table 1), which met the requirements of the qRT-PCR experiment. Therefore, these subjects reference genes (ACT, TUA, TUB, TEF, 18S rRNA, EF1α, GAPDH, UBC, YLS8, SAND, and CDPK) could serve as the research subjects. Analysis of the expression level of the 11 candidate reference genes Ct value is one of the criteria used to measure gene expression level. A gene with a higher Ct value would have a lower gene expression level [29]. Ct values of the 11 candidate reference genes were unevenly distributed in the different treatments and tissues of birch (Fig 1). How- ever, the mean Ct value can illustrate the abundance of the transcripts from the 11 candidate reference genes (Table 1), which can be used to analyze the stability of reference genes in dif- ferent experimental samples. The mean Ct values of all the samples varied from 17.39 to 29.80. Among the 11 candidate reference genes in all samples, 18S rRNA had the highest expression level, with an average Ct ± SD of 17.39 ± 1.36. ACT, TUA, TUB, TEF, EF-1α, and YLS8 had relatively high expression level, with average Ct ± SD values of 22.49 ± 1.03, 22.80 ± 1.00, 23.23 ± 1.05, 22.28 ± 1.41, 23.09 ±1.96, and 23.47 ± 1.53, respectively. CDPK had the highest average Ct value, indicating that it had the lowest expression level (Fig 1). The lower the coefficient of variation (CV) value, the higher the stability of the candidate reference gene [30]. Among the 11 candidate reference genes in all samples, the CV value of TUA was the lowest, followed by TUB and ACT. GAPDH had a high CV value, which indicated that its expression was least stable (Table 1). Thus, the results indicated that the expression levels of 11 candidate reference genes were different among the different experimental conditions. GeNorm analysis The geNorm software can identify reference genes with better stabilities by calculating the average expression stability index (M value). The software will rank candidate reference genes expression stabilities according to their M values. The geNorm algorithm reveals the appropri- ate number of reference genes required for normalization to calculate the pairwise variation between the normalization factors NFn and NFn+1 of the two sequences (Vn/n+1). When Vn/n+1 is less than 0.15, n is the number of genes used in the normalization [31–32]. The lower the M value, the more stable the reference gene. Conversely, a higher M value indicates PLOS ONE | https://doi.org/10.1371/journal.pone.0225926 December 3, 2019 5 / 13 Selection of appropriate reference genes in birch Fig 1. Cycle threshold (Ct) value distribution of eleven candidate reference genes in different tissues and abiotic stress conditions in birch. Boxes contain the 5th and 95th quartiles of the Ct values of the different genes tested, median (central horizontal line) and minimal/maximal value (vertical bar). https://doi.org/10.1371/journal.pone.0225926.g001 worse stability. Genes with an M value less than 1.5 could be used as reference genes [33]. To obtain optimum reference genes for the different experimental conditions, total samples were divided into four groups including different tissues, normal, salt, and osmosis, as described in the section describing plant material and treatments. The results are shown in Fig 2. Ten of the eleven candidate reference genes in this study could be considered as appropriate reference genes with the M value less than 1.5, except for GAPDH in salt, different tissues, and total samples. Fig 2. Expression stability and ranking of eleven candidate reference genes as analyzed by geNorm. The black spots represent the average expression stability (M). The highest M value indicates the most unstable gene, while the lowest represents the most stable. The order from left to right indicates the stability ranking of the eleven candidate reference genes. The most variable is on the left while the most stable is on the right. https://doi.org/10.1371/journal.pone.0225926.g002 PLOS ONE | https://doi.org/10.1371/journal.pone.0225926 December 3, 2019 6 / 13 Selection of appropriate reference genes in birch In the different tissues of birch, TUB and EF-1α were the most stable reference genes. In normal conditions, ACT and TUA were the most stable. Under salt stress treatment, ACT and EF-1α were optimum reference genes. Under osmotic stress treatment, ACT and TUB were the most ideal reference genes. In total samples, ACT and TUA were the most ideal choices (Fig 2). In conclusion, the most appropriate reference genes were different in the different experimental conditions. However, ACT could be the best reference gene in all experimental conditions except in different tissues; GAPDH was the least stable reference gene in all the groups in this study. NormFinder analysis NormFinder software can select the most appropriate reference gene according to candidate reference genes expression steady values [34]. A reference gene with the least steady value (SV value) has the most stable expression. NormFinder can not only compare the expression differ- ences of candidate reference genes, but also compares the variation between samples. The results of gene stability ranking of candidate reference genes were generated and are shown in Table 2. TUB was the most stable reference gene in total group and in the normal conditions group. TEF was optimum reference gene under salt and osmotic stress. GAPDH was the least stable reference gene in all experimental groups. 18S rRNA was unstable in most groups, while it was optimum in the different tissues group. The above results are basically consistent with the geNorm analysis. Bestkeeper analysis Bestkeeper software is commonly used to choose appropriate reference genes; however, the number of candidate reference genes cannot exceed 10. Thus, we chose the 10 best candidate reference genes according to the results of geNorm and NormFinder. In BestKeeper, SD and R values are used to select ideal reference genes, however, the mean standard deviation (mSD) can be used to assess whether the expression of a gene is stable [35]. A gene with an SD value less than 1.0 can be considered as ideal candidate reference genes. At the same time, the closer the R value is to 1, the better the stability of the reference gene. The results of ranking the can- didate reference genes according to their SD values are shown in Table 3. In the different tissues group, it was obvious that TEF was the best reference gene, although its R value was far from satisfactory. In normal conditions, CDPK had the lowest SD value, Table 2. Expression stability values (SV value) of candidate reference genes under different treatments of birch, as calculated using Normfinder. Rank Different tissues Normal conditions Salt stress Osmotic stress Total Gene SV Gene SV Gene SV Gene SV Gene SV 1 18S rRNA 0.038 TUB 0.356 TEF 0.272 TEF 0.225 TUB 0.302 2 EF1α 0.083 EF1α 0.416 ACT 0.276 TUB 0.289 ACT 0.340 3 SAND 0.097 SAND 0.418 TUB 0.286 EF1α 0.387 TUA 0.341 4 ACT 0.099 ACT 0.503 EF1α 0.316 ACT 0.428 YLS8 0.562 5 TUB 0.211 YLS8 0.526 TUA 0.430 YLS8 0.526 UBC 0.741 6 YLS8 0.458 TUA 0.557 UBC 0.473 SAND 0.547 SAND 0.748 7 TEF 0.483 18S rRNA 0.694 SAND 0.501 TUA 0.624 TEF 0.806 8 TUA 0.532 CDPK 0.695 18S rRNA 0.777 CDPK 0.732 EF1α 0.948 9 UBC 1.117 TEF 0.745 YLS8 0.889 UBC 0.799 CDPK 1.091 10 CDPK 1.458 UBC 0.968 CDPK 0.907 18S rRNA 1.137 18S rRNA 1.097 11 GAPDH 2.226 GAPDH 1.685 GAPDH 2.328 GAPDH 1.438 GAPDH 1.399 https://doi.org/10.1371/journal.pone.0225926.t002 PLOS ONE | https://doi.org/10.1371/journal.pone.0225926 December 3, 2019 7 / 13 Selection of appropriate reference genes in birch Table 3. Expression stability values (standard deviation (SD) value and R value) of candidate reference genes under different treatments of birch, as determined by BestKeeper. Rank Different tissues Normal conditions Salt stress Osmotic stress Total Gene SD R value Gene SD R value Gene SD R value Gene SD R value Gene SD R value 1 TEF 0.070 0.704 CDPK 0.490 0.776 TEF 0.310 0.572 TUB 0.310 0.955 ACT 0.550 0.888 2 TUA 0.090 0.414 ACT 0.630 0.912 TUA 0.350 0.753 ACT 0.350 0.773 TUA 0.550 0.891 3 TUB 0.230 0.895 18S rRNA 0.650 0.694 ACT 0.420 0.952 TUA 0.440 0.697 TUB 0.550 0.881 4 EF1α 0.250 0.997 TUA 0.710 0.909 TUB 0.420 0.869 TEF 0.480 0.950 18S rRNA 0.690 0.356 5 ACT 0.380 0.903 TUB 0.740 0.950 EF1α 0.510 0.943 YLS8 0.570 0.755 TEF 0.710 0.662 6 18S rRNA 0.570 0.941 TEF 0.950 0.901 UBC 0.510 0.750 EF1α 0.630 0.944 YLS8 0.810 0.877 7 SAND 0.630 0.993 EF1α 0.970 0.957 SAND 0.630 0.873 SAND 0.640 0.856 SAND 0.880 0.793 8 UBC 0.710 0.149 YLS8 0.980 0.914 18S rRNA 0.730 0.348 CDPK 0.710 0.630 EF1α 0.930 0.776 9 YLS8 0.780 0.997 SAND 1.020 0.947 CDPK 0.760 0.879 UBC 1.030 0.632 UBC 0.970 0.822 10 CDPK 0.780 -0.174 UBC 1.120 0.848 YLS8 0.860 0.853 18S rRNA 1.230 0.765 CDPK 1.090 0.711 https://doi.org/10.1371/journal.pone.0225926.t003 followed by ACT, and the R value of ACT was closer to 1 than that of CDPK. On the whole, CDPK was more appropriate to serve as a candidate reference gene under the normal experi- mental conditions used. Under salt stress conditions, SD values of all the candidate reference genes were less than 1.0. Thus, the 10 tested candidate reference genes could be used as refer- ence genes, however, the R values of ACT and EF1α are closer to 1, so we consider choosing them. TEF was the best candidate reference gene, which agreed with the results of NormFin- der. ACT and EF1α had R values that closer to 1. Under osmotic stress conditions, TUB was the best choice because of its ideal SD and R values. Ultimately, ACT, TUA, and TUB had simi- lar stabilities according to the results in all samples. These three candidate reference genes were suitable as candidate reference genes, which agreed with the results of geNorm. Validation of reference genes To verify the selected reference genes, the most and least stable reference genes were used to assess the relative expression level of target BpGRAS genes after normalization under salt stress and different tissues. The GRAS transcription factor family is one of the largest families of transcription factors and are involved in regulating the expression of several target function genes in plants biotic and abiotic responses, including that to salt. High-salt and drought stress could induce the expression of SCL7 in Populus euphratica, which belongs to the GRAS tran- scription factor family [36]. Eight GRAS transcription factor genes were upregulated in differ- ent tissue of Dendrobium catenatum following exposure to heat and salt stresses [37]. Our previous research showed that transiently overexpressed BpGRASs had significantly increased salt tolerance in birch (unpublished). We used the most stable and least stable candidate reference genes according to our results to normalize the expression level of BpGRAS genes. The expression patterns of the BpGRAS genes were markedly different when the most and least stable reference genes were used to nor- malize their expression in salt stress and different tissues (Fig 3). When the ideal reference genes ACT and TEF were used for normalization under treatment with 200 mM NaCl, the expression level of the BpGRAS genes were significantly enhanced compared with that in the control group at most time points, and their expression patterns were similar. However, we could not draw the same conclusion when the least stable reference gene was used (Figs 3 and 4). When the ideal reference genes ACT and TEF were used as inter- nal controls for samples treated with 200 mM NaCl, BpGRAS1 had at the highest transcript level at 48 h after salt treatment. However, when the least stable reference gene GAPDH was PLOS ONE | https://doi.org/10.1371/journal.pone.0225926 December 3, 2019 8 / 13 Selection of appropriate reference genes in birch Fig 3. Relative quantification of targeted BpGRAS gene expression levels in birch using validated reference genes. Samples of leaves collected at 0, 3, 6, 12, 24, and 48 h under salt stress, respectively. A1-C1: Samples treated with 200 mM NaCl using ACT as internal the reference gene to normalize the expression of BpGRAS genes. A2-C2: Samples treated with 200 mM NaCl using TEF as the internal reference gene to normalize the expression of BpGRAS genes. A3-C3: Samples treated with 200 mM NaCl using GAPDH as the internal reference gene to normalize the expression of BpGRAS gene. represented 0.01 < P < 0.05. https://doi.org/10.1371/journal.pone.0225926.g003 used to normalize the expression level of BpGRAS1, BpGRAS1 showed it was hardly expressed at 48 h. The results are the same as follows, when the ideal reference genes ACT and TEF were used for normalization after treatment of 200 mM NaCl, BpGRAS16 and BpGRAS19 showed higher transcript levels at 48 h after salt treatment; however, it was hardly expressed at 48 h when the least stable reference gene, GAPDH, was used for normalization. Simultaneously, T- test showed that the expression level of NaCl treatment group was significantly higher than that of the control. The most stable reference genes for use in different tissues were EF1α and TUB and the least stable reference gene was GAPDH. These three genes were used to normalize the expres- sions of the BpGRAS genes. The combination of EF1α and TUB was also used for normaliza- tion. The results are shown in Fig 4. When the appropriate reference genes or their combination were used, the expression of BpGRAS1 in root was the highest, followed by that in the stem, with the lowest expression in the leaf. The expression of BpGRAS16 was highest in the stem, followed by leaf and then the root. The expression of BpGRAS19 in the stem and root were the essentially the same. However, when the least stable reference gene was used for nor- malization, the BpGRAS genes were hardly expressed in the stem and root. Obviously, the expression patterns of the BpGRAS genes were similar after using the appropriate reference genes and their combinations for normalization. However, the expression patterns of BpGRAS genes showed large differences when the least stable reference gene was used. Fig 4. Expression levels in birch using the validated reference genes. Samples of leaves, stems, and roots in normal conditions assessed using qRT-PCR with different internal reference gene to normalize the expression of BpGRAS genes. represented 0.01 < P < 0.05. https://doi.org/10.1371/journal.pone.0225926.g004 PLOS ONE | https://doi.org/10.1371/journal.pone.0225926 December 3, 2019 9 / 13 Selection of appropriate reference genes in birch In summary, using different reference genes for normalization could produce different experimental results. Therefore, selecting the ideal reference genes according to different experimental conditions and different tissues is very importance. Discussion Determining a gene’s expression level is an important step to understand the function of the gene product in biological processes during environmental stress, and in developmental and cellular processes [38]. To ensure the accuracy of relative RT-PCR, specific PCR conditions and an appropriate internal controls must be determined. Thus, the selection of internal refer- ence genes that show stable expression in all tissues and under the experimental conditions being investigated plays a key role in determining stress-related gene expression under differ- ent abiotic stress conditions. Identifying suitable internal reference genes in birch will encour- age gene expression studies in this tree species. There are many reports concerning reference gene selection in plant, including in Salicor- nia europaea [12], cucumber [15]. However, until now, there have been no systematic reports about reference genes selection in birch. Therefore, the present study was designed to identify and validate suitable reference genes for gene expression analysis in birch. In this study, 11 candidate reference genes that have been frequently used in previous stud- ies were evaluated, and three data analysis tools (geNorm, NormFinder, BestKeeper), which are specialized for reference gene selection, were used. The expression levels of the 11 candi- date reference genes were significantly different among the different samples (Fig 1). The three analysis tools have different algorithms; therefore, the ranking of these 11 candidate reference genes according to their expression stability varied. The differences in the statistical theories used by the three pieces of software would explain the different results. Using correlation anal- ysis (R ) 11 candidate genes between M values vs SV values (ranged from 0.0022 to 0.8169), SV values vs SD values (ranged from 0.0048 to 0.8232), M values vs SD values (ranged from 0.0068 to 0.9159), which indicated the data could not establish the expression stability. We should select the reference genes with better expression stability as identified by all three tools. Therefore, the rankings of candidate reference genes calculated by these three algorithms are listed (S1 Table). In general, EF1α was the best reference gene in different tissues. ACT and TEF are the best for salt-stressed sample and TUB alone is the best for osmotic-stressed sample. Meanwhile, ACT was also the most appropriate reference gene in normal conditions and in all the samples. Previously GAPDH was identified as the optimum reference gene in salt-treated leaves, Cd-treated roots, cold-treated leaves and roots, and PEG-treated leaf samples in Carex rigescens [20]. However, GAPDH was the least stable reference gene in all the experimental conditions in the present study, despite being one of the most commonly used reference genes in previous studies. Suitable reference genes have been confirmed in many plant species in different tissues and under abiotic stress. In Salicornia europaea, ACT (Actin) and GAPDH were the optimum com- bination of internal reference genes to study gene expression under drought stress [12]. In rice, UBQ5 and eEF1α was most stable as reference genes in different tissues [39]. In tomato, the expression stabilities of GAPDH and phosphoglycerate kinase (PGK) were ranked as the top during light stress but were poorly ranked during N and cold stress [40]. In cucumber, TUA was considered as an appropriate reference gene in different tissues. EF1α was identified as a suitable reference gene for abiotic stress treatment [15]. Thus, appropriate reference genes differ among different plants and according to the experimental conditions. In conclusion, dif- ferent reference genes should be selected according to different experimental conditions to obtain accurate results. PLOS ONE | https://doi.org/10.1371/journal.pone.0225926 December 3, 2019 10 / 13 Selection of appropriate reference genes in birch Conclusion In this study, ACT was identified as the optimum reference gene in all experimental groups, except in the different tissues group. GAPDH was the least stable candidate reference gene in all experimental conditions. This study provided appropriate reference genes for expression studies in birch, which will be beneficial for more accurate relative quantification of mRNA expression in birch for different tissues, in normal conditions, and under salt and osmotic stress conditions. Supporting information S1 Fig. The results of 1.5% agarose gel electrophoresis after 30 cycles PCR. (TIF) S2 Fig. Dissociation curves of the qRT-PCR amplicons. (TIF) S1 Table. Comprehensive rankings of the stability of the reference genes by three algo- rithms. GN: Ranking of candidate reference genes calculated by geNorm. NF: Ranking of can- didate reference genes calculated by NormFinder. BK: Ranking of candidate reference genes calculated by Bestkeeper (DOCX) Author Contributions Conceptualization: Ziyi Li, Yucheng Wang, Xiaoyu Ji. Data curation: Ziyi Li. Formal analysis: Ziyi Li, Huijun Lu, Zihang He. Funding acquisition: Xiaoyu Ji. Investigation: Huijun Lu. Methodology: Xiaoyu Ji. Project administration: Chao Wang, Yucheng Wang. Software: Ziyi Li, Huijun Lu, Zihang He. Supervision: Yucheng Wang. Validation: Zihang He, Chao Wang. Visualization: Huijun Lu, Zihang He, Chao Wang. Writing – original draft: Ziyi Li, Huijun Lu. Writing – review & editing: Ziyi Li. References 1. Morandi A, Zhaxybayeva O, Gogarten JP, Graf J. Evolutionary and diagnostic implications of intrage- nomic heterogeneity in the 16S rRNA gene in Aeromonas strains. J Bacteriol. 2005; 187(18):6561– 6564. https://doi.org/10.1128/JB.187.18.6561-6564.2005 PMID: 16159790 2. Garrido-Maestu A, Chapela MJ, Vieites JM, Cabado AG. Lolb gene, a valid alternative for qPCR detec- tion of Vibrio cholerae in food and environmental samples. Food Microbiology. 2015; 46(46):535–540. PLOS ONE | https://doi.org/10.1371/journal.pone.0225926 December 3, 2019 11 / 13 Selection of appropriate reference genes in birch 3. Godhe A, Otta SK, Rehnstam-Holm AS, Karunasagar I. Polymerase chain reaction in detection of Gym- nodinium mikimotoi and Alexandrium minutum in field samples from southwest India. Mar Biotechnol (NY). 2001; 3(2):152–162. https://doi.org/10.1007/s101260000052 PMID: 14961378 4. Steeples LR, Guiver M, Jones NP. Real-time PCR using the 529 bp repeat element for the diagnosis of atypical ocular toxoplasmosis. Br J Ophthalmo. 2016; 100(2): 200–203. 5. Zhu J, Zhang L, Li W, Han S, Yang W, Qi L. Reference gene selection for quantitative real-time PCR normalization in Caragana intermedia under different abiotic stress conditions. PLoS ONE. 2013; 8: e53196. https://doi.org/10.1371/journal.pone.0053196 PMID: 23301042 6. Dean JD, Goodwin PH, Hsiang T. Comparison of relative RT-PCR and northern blot analyses to mea- sure expression ofβ-1,3-glucanase in nicotiana benthamiana, infected with colltotrichum destructivum. Plant Molecular Biology Reporter. 2002; 20(4):347–356. 7. Gare EM, Divjak M, Bailey MJ, Walters EH. Beta-Actin and GAPDH housekeeping gene expression in asthmatic airways is variable and not suitable for normalising mRNA levels. Thorax. 2002; 57(9):765– 770. https://doi.org/10.1136/thorax.57.9.765 PMID: 12200519 8. Migocka M, Papierniak A. Identification of suitable reference genes for studying gene expression in cucumber plants subjected to abiotic stress and growth regulators. Mol. Breed. 2011; 28:343–357. 9. Reddy PS, Reddy DS, Sharma KK, Bhatnagar-Mathur P, Vadez V. Cloning and validation of reference genes for normalization of gene expression studies in pearl millet by quantitative real-time PCR. Plant Gene. 2015; 1:35–42. 10. Lv S, Jiang P, Chen X, Fan P, Wang X, Li Y. Multiple compartmentalization of sodium conferred salt tol- erance in Salicornia europaea. Plant Physiol Biochem. 2011; 51:47–52. https://doi.org/10.1016/j. plaphy.2011.10.015 PMID: 22153239 11. Ma J, Zhang M, Xiao X, You J, Wang J, Wang T, et al. Global transcriptome profiling of Salicornia euro- paea L. shoots under NaCl treatment. PLoS ONE. 2013; 8:e65877. https://doi.org/10.1371/journal. pone.0065877 PMID: 23825526 12. Xiao X, Ma J, Wang J, Wu X, Li P, Yao Y. Validation of suitable reference genes for gene expression analysis in the halophyte Salicornia europaea by real-time quantitative PCR. Front Plant Sci.2014; 5:788. https://doi.org/10.3389/fpls.2014.00788 PMID: 25653658 13. Kiarash JG, Dayton WH, Amirmahani F, Mehdi MM, Zaboli M, Nazari M.Selection and validation of ref- erence genes for normalization of qRT-PCR gene expression in wheat (Triticum durum L.) under drought and salt stresses. J Genet. 2018; 97(5): 1433–1444. PMID: 30555091 14. Tang X, Wang H, Shao C, Shao H. Reference Gene Selection for qPCR Normalization of Kosteletzkya virginica under Salt Stress. Biomed Res Int. 2015: 823806. https://doi.org/10.1155/2015/823806 PMID: 15. Wan H, Zhao Z, Qian C, Sui Y, Malik AA, Chen J. Selection of appropriate reference genes for gene expression studies by quantitative real-time polymerase chain reaction in cucumber. Anal Biochem. 2010; 399(2):257–261. https://doi.org/10.1016/j.ab.2009.12.008 PMID: 20005862 16. Zhu X, Li X, Chen W, Chen J, Lu W, Chen L, et al. Evaluation of new reference genes in papaya for accurate transcript normalization under different experimental conditions. PLoS One. 2012; 7(8): e44405. https://doi.org/10.1371/journal.pone.0044405 PMID: 22952972 17. Czechowski T, Stitt M, Altmann T, Udvardi MK, Scheible WR. Genome-wide identification and testing of superior reference genes for transcript normalization in Arabidopsis. Plant Physiol.2005; 139(1):5–17. https://doi.org/10.1104/pp.105.063743 PMID: 16166256 18. Wan Q, Chen S, Shan Z, Yang Z, Chen L, Zhang C, et al. Stability evaluation of reference genes for gene expression analysis by RT-qPCR in soybean under different conditions. PLoS One.2017; 12(12): e0189405. https://doi.org/10.1371/journal.pone.0189405 PMID: 29236756 19. Shukla P, Reddy RA, Ponnuvel KM, Rohela GK, Shabnam AA, Ghosh MK, et al. Selection of suitable reference genes for quantitative real-time PCR gene expression analysis in Mulberry (Morus alba L.) under different abiotic stresses. Mol Biol Rep. 2019; 6(2):1809–1817. 20. Zhang K, Li M, Cao S, Sun Y, Long R, Kang J, et al. Selection and alidation of reference genes for target gene analysis with quantitative real-time PCR in the leaves and roots of Carex rigescens under abiotic stress. Ecotoxicol Environ Saf. 2019; 168:127–137. https://doi.org/10.1016/j.ecoenv.2018.10.049 PMID: 30384160 21. Kang D, Guo Y, Ren C, Zhao F, Feng Y, Han X, et al. Population structure and spatial pattern of main tree species in secondary Betula platyphylla forest in Ziwuling Mountains, China. Sci Rep. 2014; 4:6873. https://doi.org/10.1038/srep06873 PMID: 25362993 22. Miao L, Qin X, Gao L, Li Q, Li S, He C, et al. Selection of reference genes for quantitative real-time PCR analysis in cucumber (Cucumis sativus L.), pumpkin (Cucurbita moschata Duch.) and cucumber–pump- kin grafted plants. Peer J. 2019; 7:e6536. https://doi.org/10.7717/peerj.6536 PMID: 31024757 PLOS ONE | https://doi.org/10.1371/journal.pone.0225926 December 3, 2019 12 / 13 Selection of appropriate reference genes in birch 23. Ma R, Xu S, Zhao Y, Xia B, Wang R. Selection and Validation of Appropriate Reference Genes for Quantitative Real-Time PCR Analysis of Gene Expression in Lycoris aurea. Front Plant Sci. 2016; 7:536. https://doi.org/10.3389/fpls.2016.00536 PMID: 27200013 24. Pfaffl MW. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res. 2001; 29:e45. https://doi.org/10.1093/nar/29.9.e45 PMID: 11328886 25. Vandesompele J, De PK, Pattyn F, Poppe B, Van RN, De PA, et al. Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol. 2002; 3(7):RESEARCH0034. 26. Andersen CL, Jensen JL, Orntoft TF. 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.2004; 64:5245–5250. https://doi.org/10.1158/0008- 5472.CAN-04-0496 PMID: 15289330 27. Ramakers C, Ruijter JM, Deprez RH, Moorman AF. Assumption-free analysis of quantitative real-time polymerase chain reaction (PCR) data. Neurosci Lett. 2003; 339(1):62–66. https://doi.org/10.1016/ s0304-3940(02)01423-4 PMID: 12618301 28. Pfaffl MW, Tichopad A, Prgomet C, Neuvians TP. Determination of stable housekeeping genes, differ- entially regulated target genes and sample integrity: BestKeeper-Excel-Based tool using pair-wise cor- relations. Biotechnol Lett.2004; 26:509–515. https://doi.org/10.1023/b:bile.0000019559.84305.47 PMID: 15127793 29. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2-ΔΔCt method. Methods.2001; 25:402–408. https://doi.org/10.1006/meth.2001.1262 PMID: 30. Ni X, Qi J, Zhang G, Xu J, Tao A, Fang P, et al. Selection of reliable reference genes for quantitative real-time PCR gene expression analysis in Jute (Corchorus capsularis) under stress treatments. Front Plant Sci.2015; 6:848. https://doi.org/10.3389/fpls.2015.00848 PMID: 26528312 31. Wang JJ, Han S, Yin W, Xia X, Liu C. Comparison of Reliable Reference Genes Following Different Hor- mone Treatments by Various Algorithms for qRT-PCR Analysis of Metasequoia. International Journal of Molecular Sciences. 2018. 20(1):34. 32. Wei Y, Liu Q, Dong H, Zhou Z, Hao Y, Chen X, et al. Selection of Reference Genes for Real-Time Quan- titative PCR in Pinus massoniana.Post Nematode Inoculation. PLoS One. 2016; 11(1):e0147224. https://doi.org/10.1371/journal.pone.0147224 PMID: 26800152 33. Vandesompele J, De PK, Pattyn F, Poppe B, Van RN, De PA, et al. Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol. 2002; 3(7): RESEARCH0034. 34. Andersen CL, Jensen JL,Orntoft TF. 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.2004; 64:5245–5250. https://doi.org/10.1158/0008- 5472.CAN-04-0496 PMID: 15289330 35. Zeng X, Ling H, Chen X, Gu S. Genome-wide identification, phylogeny and function analysis of GRAS gene family in Dendrobium catenatum (Orchidaceae). Gene. 2019; 705:5–15. https://doi.org/10.1016/j. gene.2019.04.038 PMID: 30999026 36. Ma HS, Xia XL, Yin WL. Cloning and analysis of SCL7 gene from Populus euphratica. Beijing Linye Daxue Xuebao (Journal of Beijing Forestry University).2011; 33:1–10. 37. Wang Y, Liu Z, Wu Z, Li H, Wang W, Cui X, et al. Genome-wide identification and expression analysis of GRAS family transcription factors in tea plant (Camellia sinensis). Sci Rep. 2018; 8(1):3949. https:// doi.org/10.1038/s41598-018-22275-z PMID: 29500448 38. Hu R, Fan C, Li H, Zhang Q, Fu YF. Evaluation of putative reference genes for gene expression normal- ization in soybean by quantitative real-time RT-PCR. BMC Molecular Biology.2009; 10:93. https://doi. org/10.1186/1471-2199-10-93 PMID: 19785741 39. Jain M, Nijhawan A, Tyagi AK, Khurana JP. Validation of housekeeping genes as internal control for studying gene expression in rice by quantitative real-time PCR. Biochem Biophys Res Commun. 2006; 345(2):646–651. https://doi.org/10.1016/j.bbrc.2006.04.140 PMID: 16690022 40. Løvdal T, Lillo C. Reference gene selection for quantitative real-time PCR normalization in tomato sub- jected to nitrogen, cold, and light stress. Analytical Biochemistry.2009; 387(2):238–242. https://doi.org/ 10.1016/j.ab.2009.01.024 PMID: 19454243 PLOS ONE | https://doi.org/10.1371/journal.pone.0225926 December 3, 2019 13 / 13

Journal

PLoS ONE – Public Library of Science (PLoS) Journal

Published: Dec 3, 2019

There are no references for this article.