Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Selection and validation of reference genes for quantitative real-time PCR of Quercus mongolica Fisch. ex Ledeb under abiotic stresses

  • Hao Zhan,

    Roles Conceptualization, Formal analysis, Methodology, Validation, Writing – original draft

    Affiliation College of Horticulture, Shenyang Agricultural University, Shenyang, China

    ⨯
  • Hanzhang Liu,

    Roles Formal analysis

    Affiliation College of Forestry, Shenyang Agricultural University, Shenyang, China

    ⨯
  • Tianchong Wang,

    Roles Formal analysis

    Affiliation College of Forestry, Shenyang Agricultural University, Shenyang, China

    ⨯
  • Lin Liu,

    Roles Methodology

    Affiliation College of Forestry, Shenyang Agricultural University, Shenyang, China

    ⨯
  • Wanfeng Ai,

    Roles Validation

    Affiliation College of Horticulture, Shenyang Agricultural University, Shenyang, China

    ⨯
  • Xiujun Lu

    Roles Conceptualization, Funding acquisition, Writing – review & editing

    lxjsyau@syau.edu.cn

    Affiliations College of Forestry, Shenyang Agricultural University, Shenyang, China, Key Laboratory of Forest Tree Genetics, Breeding and Cultivation of Liaoning Province, Shenyang, China

    ⨯

Abstract

Quercus mongolica Fisch. ex Ledeb is the main species of coniferous and broadleaved mixed forests in northeast and north China, which has high ornamental, economic, and ecological value. The appropriate reference genes must be selected for quantitative real-time PCR to reveal the molecular mechanisms of stress responses and their contribution to breeding of Q. mongolica. In the present study, we chose 11 candidate reference genes (TUA, CYP18, HIS4, RPS13, ACT97, TUB1, UBQ10, UBC5, SAND, PP2A, and SAMDC) and used four programs (GeNorm, NormFinder, BestKeeper, and RefFinder) to assess the expression stability of the above genes in roots, stems, and leaves under five abiotic stress factors (cold, salt, drought, weak light, and heavy metal). The findings revealed that under various experimental environments, the most stable genes were different; CYP18, ACT97, and RPS13 ranked the highest under most experimental environments. Moreover, two genes induced by stress, CMO and P5CS2, were chosen to demonstrate the reliability of the selected reference genes in various tissues under various stress conditions. Our research provides a significant basis for subsequent gene function studies of Q. mongolica.

Citation: Zhan H, Liu H, Wang T, Liu L, Ai W, Lu X (2022) Selection and validation of reference genes for quantitative real-time PCR of Quercus mongolica Fisch. ex Ledeb under abiotic stresses. PLoS ONE 17(4): e0267126. https://doi.org/10.1371/journal.pone.0267126

Editor: Raffaella Balestrini, Institute for Sustainable Plant Protection, C.N.R., ITALY

Received: November 6, 2021; Accepted: April 2, 2022; Published: April 28, 2022

Copyright: © 2022 Zhan 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: The data underlying the results presented in the study are available from National Center for Biotechnology Information (https://www.ncbi.nlm.nih.gov/bioproject/PRJNA609556).

Funding: This research was funded by National Key Research and Development Program of China (Grant Numbers 2017YFD0600602). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Gene expression analysis is crucial for revealing the mechanisms underlying different biological processes, including signal transduction in plants, metabolic pathways, tissue development, and stress response at the molecular level [1–4]. qRT-PCR is presently an important strategy employed for investigating gene expression because of its specificity, accuracy, efficiency, and high sensitivity [5, 6]. However, qRT-PCR can be substantially affected by several factors such as the choice of the reference gene, amplification efficiency of reverse transcription PCR, and quality and purity of RNA. Hence, the appropriate reference gene must be chosen for correction and standardization to avoid errors in qRT-PCR assays.

The proper reference genes should be stably expressed under various experimental environments and in all sample tissues. Some housekeeping genes, such as glyceralde-hyde-3-phosphate dehydrogenase (GAPDH), actin (ACT), ubiquitin (UBQ), and beta-tubulin (TUB), have been widely employed as reference genes for plant qRT-PCR [7–10]. Nonetheless, increasing studies have revealed that the expression levels of these generally employed reference genes may differ under experimental environments or in various tissues and are generally applicable to specific plant species [11–13]. Moreover, the current research usually selects reference genes reported in model plants for gene expression analysis because of the lack of genome information regarding non-model plants [14]. Nevertheless, the stability of reference genes requires further verification because using the inappropriate reference gene will cause errors in the description of the expression level of the target gene [15]. Therefore, stable reference genes must be chosen according to the actual conditions.

The genus Quercus belongs to the Fagaceae family, containing 300 to 600 large woody deciduous and evergreen species that are distributed from tropical to temperate regions and are an important part of forests in the Northern Hemisphere [16]. Among deciduous species of the genus Quercus, Quercus mongolica Fisch. ex Ledeb is one of the dominant species with strong resistance in warm temperate forests and is the main species of coniferous and broadleaved mixed forests in northeast and north China [17]. Quercus mongolica not only has high economic value (for example, it is used to breed large-diameter timber and raise tussah, an important silk-secreting insect in China) but also plays an important role in maintaining a regional ecological balance, protecting the ecological environment, and restoring and rebuilding the ecosystem [17, 18].

Quercus mongolica has strong tolerance to cold and drought stress [18]; however, as a heliophilous species, the seedlings of Q. mongolica cannot grow normally under low light stress, which seriously affects the process of natural regeneration [19]. In addition, few reports have focused on the tolerance of Q. mongolica to other abiotic stresses. With the rapid development of molecular biology, the exploration of gene expression patterns has become possible, which can help in understanding the molecular mechanisms of stress response and can contribute to breeding and diversified utilization of Q. mongolica. Nonetheless, no study has reported the systematic verification and assessment of suitable reference genes of Q. mongolica under various abiotic stress environments; therefore, suitable reference genes for Q. mongolica stress-related gene expression analysis must be identified.

In this study, we investigated the expression stabilities for 11 candidate reference genes (TUA, CYP18, HIS4, RPS13, ACT97, TUB1, UBQ10, UBC5, SAND, PP2A, and SAMDC) in Q. mongolica leaves, stem, and roots under cold, salt, drought, weak light, and heavy metal stresses. The most stable gene for qPCR was determined using NormFinder, GeNorm, RefFinder, and BestKeeper. Finally, CMO and P5CS2 from Q. mongolica were employed to verify the suitability of the candidate reference genes. The purpose of the current work was to assess the appropriate reference genes in specific experimental environments and to lay the foundation for future research on the expression of genes of Q. mongolica.

Materials and methods

Plant materials and stress treatments

Seeds of Q. mongolica, collected from LiaoNing Province in Northeast China, were used in this study. The seeds were sown in plastic pots filled with mixture of nutrient soil, perlite, and vermiculite (v/v/v 3:1:1) and were randomly placed in the artificial growth house of Shenyang Agricultural University (41°82′N; 123°56′E), Shenyang, China. The cultivation conditions were 16 h of light and 8 h of darkness, 400 μmol/(m2·s) light intensity, and the temperatures under light and dark conditions were 25°C and 20°C, respectively. The relative humidity was 60%. After 8 weeks of growth, the seedlings that were grown consistently were randomly collected for different experimental treatments. For drought and salt stress, the seedlings were watered with 2 L 20% PEG-6000 and 2 L 300 mM NaCl, respectively. For heavy metal stress, the seedlings were watered with 2 L 1.5mM CdSO4·8/3H2O solution. For cold stress, the seedlings were subjected to temperatures of 4°C in an incubator. For weak light stress, the seedlings were placed under 80 μmol/(m2·s) light intensity. The leaves, stems, and roots were sampled separately after 24, 48, and 72 h of each treatment. Untreated seedlings of Q. mongolica at 0 h were used as control plants. All the treatments were performed in the artificial growth house mentioned above. For different tissues, the samples of leaves, stems, and roots were collected from 40-year-old Q. mongolica at Experimental Forest Farm of Liaoning Province. The experiment was performed with three biological replicates. The samples were cleaned with purified water and immediately frozen in liquid nitrogen and stored at −80°C until further analysis.

RNA isolation and cDNA synthesis

Total RNA was extracted from the leaf, stem and root tissues using a RNA Isolation Kit (TianGen, China) with DNase treatment to remove genomic DNA. 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, Waltham, MA, USA). All the RNA samples with A260/A280 ratios = 1.8–2.2 and A260/A230 ratios > 2.0 were used for further study. cDNA synthesis was performed with 1000 ng of total RNA in a final volume of 20 μL using a cDNA synthesis kit (Monad, China) following the manufacturer’s instructions. The four-fold dilution cDNA was stored at −20 ˚C for further study.

Candidate reference gene selection and primer design

According to the related research literature on plant reference genes [15, 20–23], we selected 11 candidate reference genes (TUA, CYP18, HIS4, RPS13, ACT97, TUB1, UBQ10, UBC5, SAND, PP2A, and SAMDC). We used the Arabidopsis nucleotide sequences as query sequences for BLAST in our Q. mongolica genome database (PRJNA609556) to identify candidate reference genes. The 11 reference gene sequences and homologous gene accession in The Arabidopsis Information Resource (TAIR) database were listed in S1 Table. Specific primers for the reference genes were designed using Primer premier 5.0 (Premier Biosoft International, Palo Alto, CA, USA). The parameters were as follows: the length of the primers and the amplification product size were 18–25 bp and 100–300 bp, respectively, the range of annealing temperature (Tm) was 60–70°C, and the GC was 40–60%. The primer sequences for qRT-PCR were listed in Table 1.

thumbnail
Download:
Table 1. Amplification features together with primer sequences of 11 candidate reference genes.

https://doi.org/10.1371/journal.pone.0267126.t001

To test the specificity of the qRT-PCR primers, the amplification product of PCR for each gene was verified by 1.5% agarose gel electrophoresis and the melting curve was included after amplification. Amplification efficiency (E) and correlation coefficients (R2) were calculated using a standard curve formed by using a ten-fold dilution (1, 1/10, 1/100, 1/1000, and 1/10,000) of cDNA. The amplification efficiency (E) was calculated as E = [10 (−1/slope) − 1] ×100%.

qRT-PCR analysis

The reactions were performed on a Step One Plus RT-PCR system (Applied Biosystems, USA) with TB Green Premix Ex TaqTM Ⅱ (TliRNaseH Plus) (TaKaRa, China) in accordance with the manufacturer’s instructions. Each reaction mixture (10 μL) contained 1 μL of diluted cDNA (250 ng μL−1), 5 μL of TB Green Ⅱ, 0.2 μL ROX Reference Dye, 0.5 μL of each primer (10 μM) and 2.8 μL of ddH2O. The qRT-PCR was performed under the following conditions: 95°C for 30 s, followed by 40 cycles of 95°C for 5 s, 53°C for 30 s and 72°C for 30 s, then dissolution at 95°C for 15 s, 60°C for 1 min and 95°C for 15 s. Three independent replications were performed.

Stability analysis of candidate reference genes

Three different software programs GeNorm [24], NormFinder [25], BestKeeper [26] and the RefFinder [22] web tool were used to evaluate the stability of 11 candidate reference genes under various stress conditions.

Validation of candidate reference genes

To validate the reliability of the reference genes from software programs analysis, two widely known abiotic stress-related genes, namely P5CS2 and CMO, were selected to measure the expression stability of the candidate reference genes. The gene expression levels were normalized with the two most stable, as well as the two least stable, candidate reference genes identified from this study. Finally, we used the 2-△△CT method to calculate the relative expression levels of the verified genes [12]. Statistical analyses were performed using one-way ANOVA with SPSS 21.0 software.

Results

PCR specificity and qRT-PCR amplification efficiency

Eleven candidate reference genes were selected for qRT-PCR analysis. For the above genes, Table 1 revealed the gene ID, the name of the genes, R2 and E values, and the amplicon length. We applied 1.5% Agarose gel electrophoresis to analyze the amplified products, and only one desired size band was observed; non-specific amplification or primer dimers were not observed (S1 Fig). For specific detection of PCR, after amplification, a single peak was observed in the melting curve (S2 Fig). For PCR, the value of E varied from 95.79% for TUA to 118.01% for UBC5, and for the standard curve, the correlation coefficient ranged from 0.9844 for ACT97 to 0.9997 for RPS13 (S3 Fig).

Stability of expression of candidate reference genes

The distribution of Ct values of 11 genes in all detected samples is shown in Fig 1. Overall, the mean Ct values for all samples were between 21.40 and 27.30. The Ct value in RPS13 was the lowest (21.40), whereas that of ACT97 was the highest (27.30).

thumbnail
Download:
Fig 1. The distribution of Ct values of 11 candidate reference genes of Q. mongolica leaves, stem, and roots under various abiotic stress conditions.

The line across the box and the lower and upper boundaries of the box are the 50th, 25th and 75th percentile, respectively. Whisker caps represents the maximum and minimum values.

https://doi.org/10.1371/journal.pone.0267126.g001

The Ct values of all samples were generally between 17.10 and 32.17. The Ct value of UBC5 was between 21.14 and 26.86, which means that in all samples, UBC5 is the most stable, followed by RPS13 and CYP18, with Ct values between 18.93 and 24.91 and between 23.26 and 30.96, respectively. The Ct value of SAMDC changed the most (between 18.15 and 31.50), which means that SAMDC is the most unstable reference gene in the experimental environment. Similarly, the Ct values of HIS4 and TUA were between 17.10 and 27.37 and between 18.01 and 27.78, respectively, second only to SAMDC. The above results showed that the expression levels of the reference gene were different under various stress environments and tissues. Thus, Q. mongolica reference genes must be screened under specific environments.

GeNorm analysis

In the GeNorm analysis, the M values were counted to rank the mean stability of expression in 11 candidate reference genes (S2 Table and Fig 2). Previous studies have demonstrated that M values below a threshold of 1.5 indicate appropriate reference genes. A lower value of M suggests that the expression stability of the candidate reference gene is high. In the present study, the combinations of most stable reference genes in leaves, stems, and roots samples under various abiotic stresses were different: CYP18/RPS13 in all samples, HIS4/TUB1 in different tissues (DT), CYP18/TUB1 in salt-treated leaves (SL), SAND/UBC5 in salt-treated stems (SS), CYP18/RPS13 in salt-treated roots (SR), ACT97/SAND in cold-treated leaves (CL), TUB1/HIS4 in cold-treated stem (CS), ACT97/TUB1 in cold-treated roots (CR), PP2A/UBC5 in drought-treated leaves (DL) and drought-treated roots (DR), TUB1/TUA in drought-treated stem (DS), SAND/HIS4 in Cd-treated leaves (CdL), CYP18/SAND in Cd-treated stem (CdS), CYP18/ACT97 in Cd-treated roots (CdR), CYP18/TUA in weak light-treated leaves (WL), ACT97/TUB1 in weak light-treated stem (WS), and CYP18/RPS13 in weak light-treated roots (WR).

thumbnail
Download:
Fig 2. The stability measurement (M) of the expression of 11 candidate reference genes in Q. mongolica leaves, stem, and root samples under the five abiotic stress environments analyzed by GeNorm software.

CR: Cold-treated roots; CS: Cold-treated stem; CL: Cold-treated leaves; SR: Salt-treated roots; SS: Salt-treated stem; SL: Salt-treated leaves; DR: Drought-treated roots; DS: Drought-treated stem; DL: Drought-treated leaves; CdR: Cd-treated roots; CdS: Cd-treated stem; CdL: Cd-treated leaves; WR: Weak light-treated roots; WS: Weak light-treated stem; WL: Weak light-treated leaves; DT: Different tissues; Total: All samples.

https://doi.org/10.1371/journal.pone.0267126.g002

The pairwise variation (V) between the normalization factors counted through the GeNorm algorithm can also be applied to determine the optimal number of reference genes required for an accurate normalization. The cutoff of Vn/n+1 less than 0.15 generally suggests that no additional reference genes are required for standardization. Nonetheless, a value of 0.2 was also regarded as acceptable [27]. In our study, the values of V2/3 in DT (0.062), DL (0.138), CS (0.106), CL (0.111), DR (0.099), DS (0.124), CdR (0.143), and CdL (0.086) samples were lower than 0.15, and the values of V2/3 in SR (0.194), SS (0.198), SL (0.155), CdS (0.187), CR (0.189), and WL (0.182) samples were lower than 0.2, suggesting that under these treatments, the two reference genes were sufficient to achieve exact normalization (S3 Table and Fig 3). The V3/4 values for WS (0.165) and WR (0.146) samples were lower than 0.2, indicating that three reference genes could be used for accurate standardization. The value of V8/9 in total samples (0.199) was lower than 0.2, suggesting that eight reference genes should be employed for accurate normalization. Certainly, the Vn/n+1 value of total samples was merely a recommendation and samples could be analyzed separately based on the concerned tissue and/or stress in normal set up. Nonetheless, it has been reported that the use of multiple reference genes may increase the complexity and instability of experiments [28], revealing that accurate standardization can be completed by one reference gene [29, 30]. Thus, the optimal reference gene must be chosen based on various experimental environments.

thumbnail
Download:
Fig 3. Pairwise variation (V) analysis of the 11 candidate reference genes using GeNorm under various experimental conditions.

CR: Cold-treated roots; CS: Cold-treated stem; CL: Cold-treated leaves; SR: Salt-treated roots; SS: Salt-treated stem; SL: Salt-treated leaves; DR: Drought-treated roots; DS: Drought-treated stem; DL: Drought-treated leaves; CdR: Cd-treated roots; CdS: Cd-treated stem; CdL: Cd-treated leaves; WR: Weak light-treated roots; WS: Weak light-treated stem; WL: Weak light-treated leaves; DT: Different tissues; Total: All samples.

https://doi.org/10.1371/journal.pone.0267126.g003

NormFinder analysis

NormFinder assessed the stability of the reference gene through SV; a relatively low SV value generally exhibits higher stability. NormFinder counted the SV of 11 reference genes, and the results are shown in Table 2. The expression stability of RPS13 was the highest in the total and CR samples. UBQ10 was the most stable in SL samples, whereas the most stable gene in SS samples was ACT97. In addition, TUA revealed excellent gene expression stability in the CS and SR samples, and TUB1, HIS4, and SAND revealed excellent expression stability in the WR, WS, and DS samples, respectively. PP2A displayed the highest gene expression stability in the CdL and DR samples. Simultaneously, CYP18 showed the highest gene expression stability in the WL, DL, CL, CdR, CdS, DT and all samples.

thumbnail
Download:
Table 2. Expression stability for 11 candidate reference genes counted through NormFinder.

https://doi.org/10.1371/journal.pone.0267126.t002

BestKeeper analysis

The genes with the lowest SD and CV were the most stable reference genes identified via BestKeeper; relatively low values of SD (less than 1) are generally regarded to be within an acceptable variation range. Based on the CV ± SD value produced by BestKeeper (Table 3), SAND was the most stable in WR, SL, and all samples. ACT97 seems to be the most appropriate reference gene in the CL, SR, and SS samples. CYP18 was the most stable in the DS and CS samples. HIS4 ranked the highest in the DR, WS, and WL samples. In the CdL and DL samples, UBC5 ranked the highest. TUA, UBQ10, TUB1, and PP2A were the optimal reference genes in the DT, CdR, CdS, and CR samples, respectively.

thumbnail
Download:
Table 3. Expression stability of 11 candidate reference genes counted via BestKeeper.

https://doi.org/10.1371/journal.pone.0267126.t003

RefFinder analysis

Finally, we used the RefFinder to generate an overall ranking of candidate genes. RefFinder is an online tool, which integrates the algorithms of GeNorm, NormFinder, and Bestkeeper. RefFinder assigns an appropriate weight to each gene and calculates the geometric mean of their weights for the overall final ranking based on the rankings from each program, and the smaller the value, the more stable the gene (Table 4). For the WR, WS, and total samples, two top-ranked genes were identified using the RefFinder program in accordance with the samples identified through GeNorm. Among all the samples, RPS13 and CYP18 were the most appropriate reference genes. ACT97 and HIS4 were the most stable reference genes in DT samples. SAND and UBQ10 were the most stable in SL samples, while ACT97 together with CYP18 or UBC5 were the most stable in the SR and SS samples, respectively. The most stable reference genes under cold stress were CYP18 and TUA in the CS and CL samples, while ACT97 and RPS13 were the most appropriate for the CR samples. Under drought stress, CYP18 together with UBC5 or SAND were the most stable genes in the DL and DS samples, respectively. UBC5 and PP2A were the most stable genes in the DR samples. Under heavy metal stress, PP2A and UBC5 were the most suitable in the CdL samples, and CYP18 together with ACT97 or SAND were determined the most stable in the CdR and CdS samples, respectively. Under weak light stress, CYP18 together with TUA or HIS4 were stably expressed in WL and WS samples, respectively; however, RPS13 and TUB1 were stably expressed in WR samples.

thumbnail
Download:
Table 4. Eleven candidate reference genes ranked under various conditions using RefFinder.

https://doi.org/10.1371/journal.pone.0267126.t004

Validation of the candidate reference genes

To verify the expression stability of the selected reference genes using the RefFinder software, CMO and P5CS2 gene expression levels were quantified by employing the two most unstable genes or the two most stable genes (in combination or alone) in the RefFinder ranking (Fig 4). In the SL samples, when the most stable gene and its combination were used for normalization, the P5CS2 expression trend was V-shaped between 0 and 72 h, and the expression reached the lowest point at 48 h; when standardized by the most unstable gene, the P5CS2 expression trend first decreased between 0 and 72 h, then increased, and then decreased again, whereas when UBQ10, SAND, and UBQ10 + SAND were used to calibrate the expression level of CMO, the value first increased and then decreased from 0 h to 48 h. When SAMDC was employed as a reference gene, the expression level of CMO remained stable between 0 h and 48 h and greatly decreased at 72 h. In the CS samples, when CYP18 or TUA together with their combinations were applied as the reference genes, the expression trends of CMO and P5CS2 exhibited an inverted V shape between 0 and 72 h, and there was a peak at 24 h. However, when the least stable gene UBQ10 was utilized as a reference gene, the pattern of expression in CMO and P5CS2 were completely different, and the relative expression levels of P5CS2 and CMO were abnormally increased at 24 h. In the DR samples, when PP2A and UBC5 were used as reference genes, the expression of CMO and P5CS2 first declined and then increased; however, the expression pattern evidently changed when it was normalized through the least stable gene of SAMDC, and the value increased significantly at 48 h.

thumbnail
Download:
Fig 4. Relative gene expression levels of utilizing the chosen reference genes.

The outcomes were normalized through two of the most unstable reference genes together with two most stable genes (in combination or alone) following stress treatment after 0 h, 24 h, 48 h, and 72 h. (A) The relative expression levels of P5CS2 in SL samples. (B) The relative expression levels of CMO in SL samples. (C) The relative expression levels of P5CS2 in CS samples. (D) The relative expression levels of CMO in CS samples. (E) P5CS2 and CMO expression levels in DR samples. (F) P5CS2 and CMO expression levels in DR samples. The error bar represents the standard deviation of three bio-replicates. Different letters indicate that the expression of the reference gene expression in each condition was significantly different (P < 0.05, Duncan’s test).

https://doi.org/10.1371/journal.pone.0267126.g004

Discussion

Quantitative determination of expression levels of major gene through qRT-PCR is an important method to understand gene function and reveal the response mechanism of plants under adversities [21, 22]. The accuracy of qRT-PCR is measured based on the choice of appropriate reference genes [31]. Ideally, the expression levels of reference genes should be stable under various experimental conditions [32], but such genes are virtually absent [33, 34]. Thus, the most appropriate qRT-PCR reference genes must be selected for specific experimental environments. The purpose of the current study was to identify the optimal reference genes for adoption in analyses of gene expression in Q. mongolica under various abiotic stress conditions. We evaluated 11 potential reference genes in Q. mongolica. According to the results, the most stable reference genes were not consistent across various experimental environments. For instance, RPS13 and ACT97 performed optimally in cold-treated roots, while PP2A and UBC5 were more suitable in drought-treated roots.

In our research, three programs with different algorithms, BestKeeper, NormFinder, and GeNorm, were applied to assess the expression stability of candidate genes. GeNorm and NormFinder calculate the gene expression stability and then sort the candiate reference genes by SV, whereas BestKeeper calculates the results based on the SD and CV of the Ct value [24–26]. Because of the various algorithms employed in the three software packages, they have distinct rankings for identical experimental datasets [12], despite the fact that the analysis outcomes of NormFinder and GeNorm in this study hardly changed. For instance, the analysis of BestKeeper and GeNorm suggested that PP2A and UBC5 were the two optimal reference genes in drought-treated leaves of Q. mongolica, but NormFinder indicated that SAND and CYP18 were optimal. Furthermore, the analysis of BestKeeper suggested that TUA was the optimal reference gene in different tissues of Q. mongolica, but NormFinder and GeNorm indicated that the stability of TUA was worst. To avoid biased outcomes due to the different software programs, a fourth algorithm (RefFinder) was used to analyze the reference genes. RefFinder was applied to integrate and create comprehensive candidate reference gene rankings in accordance with the geometric average of each gene weight counted using each program [35]. Many researchers also employ RankAggreg to count the ultimate ranking [8, 11, 12]. RankAggreg applies a genetic algorithm or cross-entropy Monte Carlo algorithm to calculate the ranking of reference genes [36]. The above two tools exert a crucial effect in integrating the reference gene screening results acquired from other software.

In the present study, 11 genes that are generally applied as candidate reference genes for numerous plant species were assessed. However, the outcomes of reference gene selection are not constant in various species [11, 37], tissues [38, 39], and abiotic stress environments [14, 27]. For example, our results suggested that UBQ10 was the most unstable in cold-treated stems, while excellent stability was determined in salt-treated leaves. Similarly, TUA was the most stable in cold-treated leaves, but it was unstable in salt-treated stems; ACT97 showed excellent stability in salt-treated stems and roots but was the most unstable reference gene in the Cd-treated leaves. These results suggest that reference genes exhibit tissue specificity under certain conditions [40]. This can be explained partly through the functional diversity of reference genes; the above genes are related to the components of the cytoskeleton and participate in other cellular processes under stress environments [11]. In addition, under cold stress, UBC5 revealed excellent stability in Q. mongolica leaves in the current study, but it was the most unstable reference gene under various abiotic environments in Petroselinum crispum [41]; SAND is the most stable reference gene in Q. mongolica stems under drought stress, but its stability in Festuca arundinacea leaves is poor under salt stress [7]. Similarly, SAMDC and HIS4 were the least stable reference genes under most studied abiotic stress conditions in Q. mongolica, but SAMDC and HIS4 showed excellent stability in drought-stressed leaves and salt-stressed roots of Trifolium repens L. [22] and drought-stressed leaves of Eucommia ulmoides [20], respectively. Moreover, ACT97 was most stable in different tissues of Q. mongolica and Quercus suber [23], another plants of genus Quercus, but its expression level varied significantly under various abiotic environments in our study. These results indicate that the assessment of the expression stability of the reference gene is indispensable under specific experimental conditions before analyzing the expression of the target gene.

Generally, some reference genes can be applied under a variety of experimental environments and/or tissues owing to relatively constant expression level in a specific plant [41–43]. CYP18 represents a good choice as reference gene. In our study, CYP18 was regarded as a reliable reference gene in various tissue samples under various abiotic stress environments. CYP18 has been determined as a stable reference gene for normalizing the gene expression data of different species. For instance, CYP18 is extensively applied as a reference gene with high stability in a variety of species under various stress environments [43–45]. In this study, CYP18 was found to be the most appropriate for CS, DL, CdS, CdR, and WL samples. Simultaneously, CYP18 also ranked well and was one of the best reference genes in SR, CL, DS, WS, and all samples. Therefore, CYP18 could be prioritized for qRT-PCR analysis of Q. mongolica in various environments.

We chose P5CS2 and CMO, which were the key genes of the osmoprotectant proline and glycine betaine synthesis pathways under stress conditions such as cold, salt, heat, drought, respectively, as the target genes to confirm the stability of the selected reference genes under various stress treatments and a variety of tissues [6, 27, 46, 47]. P5CS2 and CMO upregulated in other plants under most stress conditions [11, 48]. The results showed that the expression levels of the target genes were remarkably different when the most unstable and the most stable reference genes were utilized. The expression patterns of the target genes were similar when the most stable combination or reference genes was employed in the exploration under different stress conditions. However, compared with using the most stable reference gene, the target gene expression pattern changed markedly when the least stable reference gene was applied, revealing that the unsuitable reference genes may result in incorrect expression outcomes for the target genes [49]. This confirmed that the choice of appropriate reference genes is crucial for qRT-PCR analysis.

Conclusions

In our study, the stability of the expression of 11 candidate reference genes was explored in distinct tissues and under various abiotic stress conditions in Q. mongolica. According to the of RefFinder results, CYP18 was considered the optimal reference gene in cold-treated stems, Cd-treated stems, Cd-treated roots, drought-treated leaves, and weak light-treated leaves. RPS13 ranked the highest in weak light-treated roots, cold-treated roots, and all samples. PP2A and HIS4 were the most stable in drought-treated roots and weak light-treated stems, respectively. SAND, UBQ10, TUA, and UBC5 were the most appropriate in drought-treated stems and salt, cold, and Cd-treated leaves, respectively. ACT97 seemed to be the most appropriate in different tissues, salt-treated stems and roots. The data confirmed expression patterns of the different reference genes under various experimental environments, and we focused on the significance of choosing appropriate reference genes for qRT-PCR analysis in Q. mongolica. These outcomes will provide a significant basis for investigating the molecular mechanisms underlying the abiotic stress response of Q. mongolica.

Supporting information

S1 Fig. Specific PCR amplification products of 11 reference genes. M, DL2000 DNA marker.

https://doi.org/10.1371/journal.pone.0267126.s001

(TIF)

S2 Fig. Melting curves for the 11 candidate reference genes in Q. mongolica by qRT-PCR.

https://doi.org/10.1371/journal.pone.0267126.s002

(TIF)

S3 Fig. Standard curves of 11 candidate reference genes.

https://doi.org/10.1371/journal.pone.0267126.s003

(PDF)

S1 Table. The annotation and comparison information of the 11 reference genes.

https://doi.org/10.1371/journal.pone.0267126.s004

(XLSX)

S2 Table. Gene expression stability (M) of candidate reference genes calculated by geNorm.

https://doi.org/10.1371/journal.pone.0267126.s005

(DOCX)

S3 Table. Pairwise variation (Vn/n+1) analysis of eleven candidate reference genes calculated by geNorm.

https://doi.org/10.1371/journal.pone.0267126.s006

(DOCX)

Acknowledgments

We would like to thank Editage (www.editage.cn) for English language editing.

References

  1. 1. Ji Y, Nunez Ocana D, Choe D, Larsen DH, Marcelis LFM, Heuvelink E. Far-red radiation stimulates dry mass partitioning to fruits by increasing fruit sink strength in tomato. New Phytol. 2020; 228(6):1914–1925. http://doi.org/10.1111/nph.16805 pmid:32654143
  2. 2. Bustin SA, Benes V, Nolan T, Pfaffl MW. Quantitative real-time RT-PCR—a perspective. J Mol Endocrinol. 2005; 34(3):597–601. http://doi.org/10.1677/jme.1.01755 pmid:15956331
  3. 3. Heng Y, Jiang Y, Zhao X, Zhou H, Wang X, Deng XW, et al. BBX4, a phyB-interacting and modulated regulator, directly interacts with PIF3 to fine tune red light-mediated photomorphogenesis. Proc Natl Acad Sci U S A. 2019; 116(51):26049–26056. http://doi.org/10.1073/pnas.1915149116 pmid:31776262
  4. 4. Wang F, Chen X, Dong S, Jiang X, Wang L, Yu J, et al. Crosstalk of PIF4 and DELLA modulates CBF transcript and hormone homeostasis in cold response in tomato. Plant Biotechnol J. 2020; 18(4):1041–1055. http://doi.org/10.1111/pbi.13272 pmid:31584235
  5. 5. 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. http://doi.org/10.1104/pp.105.063743 pmid:16166256
  6. 6. Kim SJ, Na JU, Kim JS, Lee JE, Nie H, Lee KA, et al. Identification of valid reference genes for quantitative RT-PCR in Caragana microphylla under salt and drought stresses. Physiol Mol Biol Plants. 2020; 26(10):2103–2108. http://doi.org/10.1007/s12298-020-00874-3 pmid:33088054
  7. 7. Yang Z, Chen Y, Hu B, Tan Z, Huang B. Identification and validation of reference genes for quantification of target gene expression with quantitative real-time PCR for tall fescue under four abiotic stresses. PLoS One. 2015; 10(3):e0119569. http://doi.org/10.1371/journal.pone.0119569 pmid:25786207
  8. 8. Li L, Li N, Fang H, Qi X, Zhou Y. Selection and Validation of Reference Genes for Normalisation of Gene Expression in Glehnia littoralis. Sci Rep. 2020; 10(1):7374. http://doi.org/10.1038/s41598-020-63917-5 pmid:32355237
  9. 9. Galli V, Borowski JM, Perin EC, Messias Rda S, Labonde J, Pereira Idos S, et al. Validation of reference genes for accurate normalization of gene expression for real time-quantitative PCR in strawberry fruits using different cultivars and osmotic stresses. Gene. 2015; 554(2):205–214. http://doi.org/10.1016/j.gene.2014.10.049 pmid:25445290
  10. 10. Kanakachari M, Solanke AU, Prabhakaran N, Ahmad I, Dhandapani G, Jayabalan N, et al. Evaluation of Suitable Reference Genes for Normalization of qPCR Gene Expression Studies in Brinjal (Solanum melongena L.) During Fruit Developmental Stages. Appl Biochem Biotechnol. 2016; 178(3):433–450. pmid:26472671
  11. 11. Zhang K, Li M, Cao S, Sun Y, Long R, Kang J, et al. Selection and validation 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. http://doi.org/10.1016/j.ecoenv.2018.10.049 pmid:30384160
  12. 12. Song H, Mao W, Duan Z, Que Q, Zhou W, Chen X, et al. Selection and validation of reference genes for measuring gene expression in Toona ciliata under different experimental conditions by quantitative real-time PCR analysis. BMC Plant Biol. 2020; 20(1):450. http://doi.org/10.1186/s12870-020-02670-3 pmid:33003996
  13. 13. Shivhare R, Lata C. Selection of suitable reference genes for assessing gene expression in pearl millet under different abiotic stresses and their combinations. Sci Rep. 2016; 6:23036. http://doi.org/10.1038/srep23036 pmid:26972345
  14. 14. Lv Y, Li Y, Liu X, Xu K. Identification of Ginger (Zingiber officinale Roscoe) Reference Genes for Gene Expression Analysis. Front Genet. 2020; 11:586098. http://doi.org/10.3389/fgene.2020.586098 pmid:33240331
  15. 15. Niu L, Tao YB, Chen MS, Fu Q, Li C, Dong Y, et al. Selection of Reliable Reference Genes for Gene Expression Studies of a Promising Oilseed Crop, Plukenetia volubilis, by Real-Time Quantitative PCR. Int J Mol Sci. 2015; 16(6):12513–12530. http://doi.org/10.3390/ijms160612513 pmid:26047338
  16. 16. Torre S, Tattini M, Brunetti C, Fineschi S, Fini A, Ferrini F, et al. RNA-seq analysis of Quercus pubescens Leaves: de novo transcriptome assembly, annotation and functional markers development. PLoS One. 2014; 9(11):e112487. http://doi.org/10.1371/journal.pone.0112487 pmid:25393112
  17. 17. Wang G, Wang Y, Qin L. Transcriptome Sequencing and Gene Function Annotation of Quercus mongolica Leaf. Science of Sericulture. 2020; 46(5):560–565. http://doi.org/10.13441/j.cnki.cykx.2020.05.005
  18. 18. Jiang Y, Liu W, Wang G, Zhou X, Qin L. Research Advances in Germplasm Resource and Utilization of Quercus L. Science of Sericulture. 2019; 45(4):577–585. http://doi.org/10.13441/j.cnki.cykx.2019.04.016
  19. 19. Zhang M, Zhu J, Li M, Zhang G, Yan Q. Different light acclimation strategies of two coexisting tree species seedlings in a temperate secondary forest along five natural light levels. Forest Ecology and Management. 2013; 306:234–242. http://doi.org/10.1016/j.foreco.2013.06.031
  20. 20. Ye J, Jin CF, Li N, Liu MH, Fei ZX, Dong LZ, et al. Selection of suitable reference genes for qRT-PCR normalisation under different experimental conditions in Eucommia ulmoides Oliv. Sci Rep. 2018; 8(1):15043. http://doi.org/10.1038/s41598-018-33342-w pmid:30301911
  21. 21. Li Z, Lu H, He Z, Wang C, Wang Y, Ji X. Selection of appropriate reference genes for quantitative real-time reverse transcription PCR in Betula platyphylla under salt and osmotic stress conditions. PLoS One. 2019; 14(12):e0225926. http://doi.org/10.1371/journal.pone.0225926 pmid:31794584
  22. 22. Pu Q, Li Z, Nie G, Zhou J, Liu L, Peng Y. Selection and Validation of Reference Genes for Quantitative Real-Time PCR in White Clover (Trifolium repens L.) Involved in Five Abiotic Stresses. Plants (Basel). 2020; 9(8). http://doi.org/10.3390/plants9080996 pmid:32764378
  23. 23. Marum L, Miguel A, Ricardo CP, Miguel C. Reference gene selection for quantitative real-time PCR normalization in Quercus suber. PLoS One. 2012; 7(4):e35113. http://doi.org/10.1371/journal.pone.0035113 pmid:22529976
  24. 24. Vandesompele J, Preter KD, Pattyn F, Poppe B, Roy NV, Paepe AD, et al. Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol. 2002; 3(7):research0034.1. http://doi.org/10.1186/gb-2002-3-7-research0034 pmid:12184808
  25. 25. 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(15):5245–5250. http://doi.org/10.1158/0008-5472.CAN-04-0496 pmid:15289330
  26. 26. Pfaffl MW, Tichopad A, Prgomet C, Neuvians TP. Determination of stable housekeeping genes, differentially regulated target genes and sample integrity: BestKeeper–Excel-based tool using pair-wise correlations. Biotechnol. Lett. 2004; 26(6):509–515. http://doi.org/10.1023/b:bile.0000019559.84305.47 pmid:15127793
  27. 27. Zhao D, Wang X, Chen J, Huang Z, Huo H, Jiang C, et al. Selection of reference genes for qPCR normalization in buffalobur (Solanum rostratum Dunal). Sci Rep. 2019; 9(1):6948. http://doi.org/10.1038/s41598-019-43438-6 pmid:31061419
  28. 28. Zhang Y, Peng X, Liu Y, Li Y, Luo Y, Wang X, et al. Evaluation of suitable reference genes for qRT-PCR normalization in strawberry (Fragaria x ananassa) under different experimental conditions. BMC Mol Biol. 2018; 19(1):8. http://doi.org/10.1186/s12867-018-0109-4 pmid:29933763
  29. 29. Wu J, Zhang H, Liu L, Li W, Wei Y, Shi S. Validation of Reference Genes for RT-qPCR Studies of Gene Expression in Preharvest and Postharvest Longan Fruits under Different Experimental Conditions. Front Plant Sci. 2016; 7:780. http://doi.org/10.3389/fpls.2016.00780 pmid:27375640
  30. 30. Xu L, Xu H, Cao Y, Yang P, Feng Y, Tang Y, et al. Validation of Reference Genes for Quantitative Real-Time PCR during Bicolor Tepal Development in Asiatic Hybrid Lilies (Lilium spp.). Front Plant Sci. 2017; 8:669. http://doi.org/10.3389/fpls.2017.00669 pmid:28487721
  31. 31. Kotrade P, Sehr EM, Wischnitzki E, Brüggemann W. Comparative transcriptomics-based selection of suitable reference genes for normalization of RT-qPCR experiments in drought-stressed leaves of three European Quercus species. Tree Genetics & Genomes. 2019; 15(3). http://doi.org/10.1007/s11295-019-1347-4
  32. 32. Huggett J, Dheda K, Bustin S, Zumla A. Real-time RT-PCR normalisation; strategies and considerations. Genes Immun. 2005; 6(4):279–284. http://doi.org/10.1038/sj.gene.6364190 pmid:15815687
  33. 33. Artico S, Nardeli SM, Brilhante O, Grossi-de-Sa MF, Alves-Ferreira M. Identification and evaluation of new reference genes in Gossypium hirsutum for accurate normalization of real-time quantitative RT-PCR data. BMC Plant Biol. 2010; 10:49. http://doi.org/10.1186/1471-2229-10-49 pmid:20302670
  34. 34. Chen X, Mao Y, Huang S, Ni J, Lu W, Hou J, et al. Selection of Suitable Reference Genes for Quantitative Real-time PCR in Sapium sebiferum. Front Plant Sci. 2017; 8:637. http://doi.org/10.3389/fpls.2017.00637 pmid:28523004
  35. 35. Gao D, Kong F, Sun P, Bi G, Mao Y. Transcriptome-wide identification of optimal reference genes for expression analysis of Pyropia yezoensis responses to abiotic stress. BMC Genomics. 2018; 19(1):251. http://doi.org/10.1186/s12864-018-4643-8 pmid:29653512
  36. 36. Pihur V, Datta S, Datta S. Weighted rank aggregation of cluster validation measures: a Monte Carlo cross-entropy approach. Bioinformatics. 2007; 23(13):1607–1615. http://doi.org/10.1093/bioinformatics/btm158 pmid:17483500
  37. 37. Zhang S, Zeng Y, Yi X, Zhang Y. Selection of suitable reference genes for quantitative RT-PCR normalization in the halophyte Halostachys caspica under salt and drought stress. Sci Rep. 2016; 6:30363. http://doi.org/10.1038/srep30363 pmid:27527518
  38. 38. Li C, Hu L, Wang X, Liu H, Tian H, Wang J. Selection of reliable reference genes for gene expression analysis in seeds at different developmental stages and across various tissues in Paeonia ostii. Mol Biol Rep. 2019; 46(6):6003–6011. http://doi.org/10.1007/s11033-019-05036-7 pmid:31446531
  39. 39. Zhang MF, Liu Q, Jia GX. Reference gene selection for gene expression studies in lily using quantitative real-time PCR. Genet Mol Res. 2016; 15(2). http://doi.org/10.4238/gmr.15027982
  40. 40. Deng Y, Li Y, Sun H. Selection of reference genes for RT-qPCR normalization in blueberry (Vaccinium corymbosum x angustifolium) under various abiotic stresses. FEBS Open Bio. 2020; 10(8):1418–1435. http://doi.org/10.1002/2211-5463.12903 pmid:32438500
  41. 41. Li MY, Song X, Wang F, Xiong AS. Suitable Reference Genes for Accurate Gene Expression Analysis in Parsley (Petroselinum crispum) for Abiotic Stresses and Hormone Stimuli. Front Plant Sci. 2016; 7:1481. http://doi.org/10.3389/fpls.2016.01481 pmid:27746803
  42. 42. Zhao J, Zhou M, Meng Y. Identification and Validation of Reference Genes for RT-qPCR Analysis in Switchgrass under Heavy Metal Stresses. Genes (Basel). 2020; 11(5). http://doi.org/10.3390/genes11050502 pmid:32375288
  43. 43. Borges AF, Fonseca C, Ferreira RB, Lourenco AM, Monteiro S. Reference gene validation for quantitative RT-PCR during biotic and abiotic stresses in Vitis vinifera. PLoS One. 2014; 9(10):e111399. http://doi.org/10.1371/journal.pone.0111399 pmid:25340748
  44. 44. Li L, Wang K, Zhao M, Li S, Jiang Y, Zhu L, et al. Selection and validation of reference genes desirable for gene expression analysis by qRT-PCR in MeJA-treated ginseng hairy roots. PLoS One. 2019; 14(12):e0226168. http://doi.org/10.1371/journal.pone.0226168 pmid:31805178
  45. 45. Qu R, Miao Y, Cui Y, Cao Y, Zhou Y, Tang X, et al. Selection of reference genes for the quantitative real-time PCR normalization of gene expression in Isatis indigotica fortune. BMC Mol Biol. 2019; 20(1):9. http://doi.org/10.1186/s12867-019-0126-y pmid:30909859
  46. 46. Yamada N, Takahashi H, Kitou K, Sahashi K, Tamagake H, Tanaka Y, et al. Suppressed expression of choline monooxygenase in sugar beet on the accumulation of glycine betaine. Plant Physiol Biochem. 2015; 96:217–221. http://doi.org/10.1016/j.plaphy.2015.06.014 pmid:26302482
  47. 47. Annunziata MG, Ciarmiello LF, Woodrow P, Dell’Aversana E, Carillo P. Spatial and Temporal Profile of Glycine Betaine Accumulation in Plants Under Abiotic Stresses. Front Plant Sci. 2019; 10:230. http://doi.org/10.3389/fpls.2019.00230 pmid:30899269
  48. 48. Xu Z, Sun M, Jiang X, Sun H, Dang X, Cong H, Qiao F. Glycinebetaine Biosynthesis in Response to Osmotic Stress Depends on Jasmonate Signaling in Watermelon Suspension Cells. Front Plant Sci. 2018; 9:1469. http://doi.org/10.3389/fpls.2018.01469 pmid:30369936
  49. 49. Ambroise V, Legay S, Guerriero G, Hausman JF, Cuypers A, Sergeant K. Selection of Appropriate Reference Genes for Gene Expression Analysis under Abiotic Stresses in Salix viminalis. Int J Mol Sci. 2019; 20(17):4210. http://doi.org/10.3390/ijms20174210 pmid:31466254