1. Introduction
Tree peony (
Paeonia suffruticosa Andrews.) is a perennial woody plant, which belongs to Paeoniaceae,
Paeonia, section Moutan. As a kind of multi-purpose plant with economic value, tree peony has high ornamental value because of its rich varieties, gorgeous colors, and diverse flowers [
1,
2]. In addition, it also has high nutritional and medicinal values because its root can be used as medicine and the seeds can be used to extract oil, which is beneficial to human health [
3,
4,
5,
6]. The natural flowering period of this tree is around April, which is a one-flower-a-year plant, and a few varieties have the phenomenon of second flowering [
7,
8]. Among them, middle-flowering tree peony varieties dominate, and early-flowering and late-flowering varieties are few. The characteristics of a relatively short and concentrated flowering period seriously affect the ornamental value of tree peony and limit the development of tree peony industry. Therefore, understanding the internal molecular mechanism of tree peony flowering may be helpful to discover new genes related to flowering traits, which can be used to regulate the flowering time of tree peony, prolong the group flowering period, and increase economic benefits.
At present, the research on tree peony mainly focuses on the construction of tissue culture rapid propagation system [
9], cultivation and domestication [
10], genetic diversity [
11], seed oil analysis [
12], and gene cloning [
13]. In recent years, with the rapid development of biotechnology, transcriptome analysis [
14], miRNA identification [
15], and other related molecular biology studies have been gradually carried out, and gene expression correlation studies have gradually become the focus of research.
Gene expression analysis is an important tool to elucidate complex regulatory processes, such as genetics, signal transduction, and metabolic pathways in the plant life cycle [
16]. Real-time quantification PCR can be used to verify gene expression levels and the reliability of sequencing results [
17,
18]. Compared with conventional real-time PCR (RT-PCR), quantitative real-time PCR (qRT-PCR) has the characteristics of high sensitivity, strong specificity, and low cost and has been widely used in gene expression level analysis [
19,
20]. However, the stability and accuracy of qRT-PCR results are affected by many factors, among which the reference gene plays a role in the correction and standardization of the results and is one of the important factors affecting the reliability of qRT-PCR results. Transcription levels of reference genes may change in different species, tissues, treatments, and developmental stages [
21]. Therefore, it is necessary to select the most suitable reference gene according to the test requirements. At present, some software and algorithms for evaluating the stability of reference genes have been developed, such as geNorm, NormFinder, and Bestkeeper, and these algorithms have been successfully used to screen out the reference genes related to biotic and abiotic stress, growth, and development in plants, such as
Zea mays [
22],
Hylocereus sp. [
23],
Allium sativum [
24], and Itoh peony [
25].
With the application of high-throughput sequencing and other technologies, a large number of microRNAs (miRNAs) have been found in many higher plants. miRNA is a class of small single-stranded non-coding RNA molecules with a length of about 20–24 nucleotides that regulate gene expression at the post-transcriptional level through sequence complementation [
26]. The first plant miRNA was first discovered in
Arabidopsis in 2002 [
27]. Studies have shown that miRNAs have been reported to exist widely in a variety of plants, and miRNAs are involved in regulating plant growth and development, morphogenesis, and environmental stress response, such as drought, salt, and temperature [
28,
29]. It has been found that miRNAs play an important role in the flowering mechanism by regulating the expression of flowering genes [
30]. For example, miR172 induced rice flowering by inhibiting the expression of
OsIDS1 and
SNB of the target gene AP2 family [
31]. miR159 could delay the flowering time by reducing the expression of the target gene MYB transcription factor in
Arabidopsis [
32]. Overexpression of miR171 could inhibit the expression of the target gene
Scarecrow-like (
SCL), which in turn leads to a late-flowering phenotype in Hordeum vulgare transgenic plants, accompanied by a smaller and translucent anther phenotype [
33]. Therefore, it is also of great significance to explore miRNA related to tree peony flowering so as to realize the regulation of tree peony flowering. However, there was only one report on reference miRNA in tree peony for different bud development processes, flower development processes, and different tissues [
34], but the screening of reference miRNA among different early- and late-flowering tree peony varieties has not been reported.
In this study, the petals of 42 different early- and late-flowering tree peony varieties were used as experimental materials. Primers of U6 (snRNA) and 15 miRNAs from small RNA sequencing were designed, resulting in a total of 16 candidate genes selected as candidate reference genes for miRNA expression normalization. qRT-PCR was used to detect and analyze the expression stability of 16 candidate reference miRNAs in 42 different tree peony varieties. The stability of 16 candidate genes was evaluated by geNorm, NormFinder, Bestkeeper, and RefFinder, and the optimal reference miRNA was verified. The results of this study will provide reference miRNAs for the expression of other target miRNAs in different tree peony varieties.
3. Discussion
Because the differences between species, materials, and tissues will affect the stability and reliability of quantitative results, it is necessary to select appropriate reference genes for data standardization during qRT-PCR. In this study, 42 early- and late-flowering tree peony varieties were used as materials, and the best reference miRNA was successfully screened out, which provided suitable reference miRNA for miRNA research of different tree peony varieties.
In general, the same reference does not have stability all the time. Under drought stress, the most suitable reference miRNAs in roots and leaves of
Glycine max were miR156a and miR167a [
35]. In
Allium sativum, the most stable reference miRNA for different explants was
AsmiR168a-5p and for different genotypes, the most stable reference miRNA was
AsmiR159a-1 [
36]. In
Juglans regia, the most suitable reference miRNAs for flower buds at different differentiation stages were
jre-miR394a,
jre-miR159a, and
jre-miR159c, and the most suitable reference miRNAs for leaf buds at different differentiation stages were 5.8S rRNA and
jre-miRn3 [
37]. The most stable reference miRNA combinations during seed development in
Brassica napus were
miR167-1_2,
miR11-1, and
miR159-1 [
38].
In this study, the stability analysis of 16 candidate reference miRNAs in 42 different tree peony varieties showed that
PsPC-5p-19095 and
PsPC-3p-51259 had the highest stability in the geNorm software.
PsPC-3p-6660 had the highest stability in the NormFinder software.
PsMIR319-p5 has the highest stability in the Bestkeeper software. In the results of comprehensive evaluation and analysis using RefFinder, the candidate reference miRNA with the highest stability was
PsPC-3p-6660, which was consistent with the results of NormFinder and slightly different from the analysis results of geNorm and Bestkeeper software. The difference between the analysis results of each software has also appeared in the previous research results [
39]. The reason for the difference in the stability of each candidate reference miRNA may be due to the large number of candidate reference miRNAs selected. Secondly, the difference in mathematical algorithms between different software will also affect the stability ranking of the test results to a certain extent. The results of several software showed that the stability of
PsmiR171k-3p was the lowest, indicating that the candidate miRNA was not suitable as the reference gene in different tree peony varieties.
At present,
U6 is one of the most common reference genes, which is used as the reference gene in miRNA quantitative expression analysis of various plants, such as
Vitis vinifera [
40],
Brassica oleracea [
41], and
Jatropha curcas [
42]. Studies have shown that
U6 is not stable in all cases [
43].
U6 is a suitable reference miRNA for different tissues and stem tissues under drought stress in
Hylocereus polyrhizus [
44]. However, in the evaluation of the expression stability of walnut flower bud and leaf bud differentiation, tissue parts, and varieties,
U6 had the worst stability and was not suitable as the reference miRNA. This phenomenon also exists in plants, such as wheat and longan [
45]. In this study,
U6 was used as a candidate reference miRNA, and its stability was low in the four software, which was not suitable as the reference miRNA for different tree peony varieties.
4. Materials and Methods
4.1. Plant Materials
A total of 42, i.e., 21 early-flowering and 21 late-flowering, tree peony varieties were selected from the experimental farm of Henan University of Science and Technology (112°24′52.05″ E, 34°35′45.91″ N) as experimental material for the evaluation of the expression stability of candidate reference genes from April to May 2021. The samples were frozen with liquid nitrogen and then stored in a −80 °C refrigerator for later use (
Table 4).
4.2. Primers Design of Candidate Reference miRNAs
In this study, U6 and 15 miRNAs with relatively stable expression levels were selected from the small RNA sequencing data of Paeonia ostii ‘Fengdan’, Mutant plants of Paeonia ostii ‘Fengdan’, and Paeonia suffruticosa ‘Lianhe’ (CNGBdb, accession number CNP0002984) at the blooming stage (BS), initial flowering stage (IF), full blooming stage (FB), and decay stage (DE) in our laboratory, including 6 known miRNAs (PsmiR159a, PsmiR858-3p, PsMIR11609-p5, PsmiR171k-3p, PsmiR11607, and PsMIR319-p5) and 9 novel miRNAs (PsPC-3p-70893, PsPC-3p-18408, PsPC-5p-19095, PsPC-3p-51259, PsPC-3p-23386, PsPC-3p-13662, PsPC-3p-15676, PsPC-3p-6660, and PsPC-5p-9292).
qRT-PCR primers were designed using poly (A) by Primer Premier 5.0 software. Primers should avoid special structures such as dimers and hairpin structures to prevent adverse effects on test results. Then, they were sent to Sangon Biotech (Shanghai) Co., Ltd. (Shanghai, China) for synthesis (
Table 5).
4.3. Isolation of miRNA and Synthesis of cDNA
miRNAs were isolated from the petals of 42 tree peony varieties using the miRcute Plant miRNA Isolation Kit (Tiangen, Beijing, China). Then, miRcute Plus miRNA First-Strand cDNA Synthesis Kit (Tiangen, China) was used to synthesize the first strand. Reaction system (20 μL): Total miRNA 2 μL, 2× miRNA RT Reaction Buffer 10 μL, miRNA RT Enzyme Mix 2 μL, and RNase-free ddH2O 6 μL. The reaction procedure was as follows: 42 °C for 60 min and 95 °C for 3 min.
4.4. Primers Specificity Analysis
Using cDNA as a template, qRT-PCR primers were used for PCR amplification using a 2× PCR Taq Master Mix (Blue Dye) (Nobelab Biotech, Beijing, China). Reaction system (25 μL): 2× PCR Taq Master Mix 12.5 μL, Primer F (10 μM) 1 μL, Primer R (10 μM) 1 μL, cDNA 2 μL, ddH2O 8.5 μL. The reaction procedure was as follows: 95 °C for 2 min; 95 °C 30 s, 60 °C 30 s, 72 °C 10 s, 35 cycles; 72 °C for 2 min.
cDNAs were diluted in a gradient of 51, 52, 53, 54, and 55. An SYBR® Green Premix Pro Taq HS qPCR Kit (Accurate Biology, Changsha, China) was used for qRT-PCR. Reaction system (20 μL): 2× SYBR® Green Pro Taq HS Premix 10 μL, Primer F (10 μM) 0.4 μL, Universal Primer R (10 μM) 0.4 μL, cDNA 2 μL, ddH2O 7.2 μL. The reaction procedure was as follows: 95 °C 30 s; 95 °C 5 s, 60 °C 30 s, 40 cycles. Three technique replicates were set for each sample.
4.5. Expression Stability Analysis of 16 Candidate Reference miRNAs
The stability of 16 candidate reference miRNAs was evaluated and ranked using geNorm, NormFinder, and Bestkeeper software, respectively. And then the online tool RefFinder (
http://blooge.cn/RefFinder/, accessed on 13 February 2023) [
46,
47] was used to comprehensively evaluate and analyze the results obtained by the above three software. Finally, the most suitable reference miRNA among different tree peony varieties was screened.
4.6. Validation of Candidate Reference miRNAs
The most stable and the unstable miRNAs were selected as the reference miRNAs, and the expression patterns of
PomiR171 and
PomiR414 (CNGBdb, accession number CNP0002984) in response to different flowering times of tree peony in different tree peony varieties were analyzed. Three technique replicates were set for each sample, and the qRT-PCR primers were given in
Table 6. The expression levels of the miRNA were calculated using the formula of 2
−ΔΔCt, and statistical analysis and mapping were performed using SPSS 22.1 (one-way ANOVA, Duncan’s test,
p < 0.05) and Origin 2018.
5. Conclusions
Using petals of 42 different early- and late-flowering tree peony varieties as experimental materials, the expression stability of 16 candidate reference genes was evaluated and analyzed by geNorm, NormFinder, Bestkeeper, and RefFinder software using qRT-PCR. The results showed that the average Ct values of all candidate reference miRNAs were between 15.34 ± 0.29 and 32.64 ± 0.38. In the geNorm software, PsPC-5p-19095 and PsPC-3p-51259 had the highest stability, and the optimal number of reference miRNAs was four, which were PsPC-5p-19095, PsPC-3p-51259, PsmiR159a, and PsPC-3p-6660. The stability of PsPC-3p-6660 was the highest in the analysis results of the NormFinder software. Among the analysis results of the Bestkeeper software, PsMIR319-p5 had the highest stability. Among the results of comprehensive evaluation and analysis of several software using RefFinder, the candidate reference miRNA with the highest stability was PsPC-3p-6660. Therefore, PsPC-3p-6660 can be used as the reference miRNA for miRNA studies of different tree peony varieties. This provides a reference miRNA for the study of other miRNAs in different tree peony varieties.