Next Article in Journal
Xp22.33 Duplication Encompassing PAR1 in a Male with Syndromic Neurodevelopmental Disorder and Tall Stature
Previous Article in Journal
Environmental Substances Associated with Neurodegeneration: An Overview of Parkinson’s Disease and Related Genotoxic Endpoints
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Genes
Volume 17
Issue 2
10.3390/genes17020237
genes-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Transcriptome-Based Selection and Validation of Reference Genes for Gene Expression Analysis in Roegneria ciliaris ‘Liao Sheng’ Across Various Tissues and Under Drought Stress

by
Qianyun Luo
,
Yue Liu
,
Yifan Wang
,
Guanghao Zhang
,
Jiafen Liu
,
Hongxin Li
,
Zhen Liang
,
Ying Liu
,
Long Bai
and
Sijia Liu
*
Department of Grassland Science, College of Horticulture, Shenyang Agricultural University, Shenyang 110866, China
*
Author to whom correspondence should be addressed.
Genes 2026, 17(2), 237; https://doi.org/10.3390/genes17020237
Submission received: 19 January 2026 / Revised: 9 February 2026 / Accepted: 12 February 2026 / Published: 14 February 2026
(This article belongs to the Section Plant Genetics and Genomics)
Browse Figures
Versions Notes

Abstract

Backgrounds: Roegneria ciliaris is a perennial tetraploid wild relative of wheat that is widely distributed in China. It can be used both as a forage crop and ecological grass (the grasses specifically bred for ecological restoration) due to its strong stress tolerance, early green-up, vigorous seedling growth in spring, and great palatability. Methods: It is necessary to select and validate appropriate reference genes (RGs) for gene expression normalization by qRT-PCR in order to decipher the stress tolerance mechanism of this grass species. Therefore, eight candidate RGs were identified from transcriptome data of R. ciliaris ‘Liao sheng’ in response to drought stress. The expression stability of these RGs was evaluated by five algorithms (∆Ct, geNorm, NormFinder, Bestkeeper and ReFinder) using samples from different tissues and drought stress. Results: The results showed that MDH and RPL19 were the most stable RGs among all samples, while GAPDH and TUBA presented the lowest expression stability. These representative RGs were further used to normalize the expression level of the pyrroline-5-carboxylate synthase (P5CS) and protein phosphatase 2C (PP2C) genes in different tissues and under drought stress. The results of P5CS and PP2C expression were consistent with transcriptome data. Conclusion: Our study provided the first systematic evaluation of the most stable RG selection for qRT-PCR normalization in R. ciliaris, which will promote further research on its tissue-specific gene expression and mechanism of drought tolerance.

1. Introduction

Roegneria ciliaris (Trin.) Nevski (2n = 4x = 28, genome ScScYcYc) is a perennial herbaceous species widely distributed mainly throughout China. It represents a valuable wild grass with multiple advantageous traits, including strong stress tolerance, early green-up and fast growth in spring, utility for windbreak and sand fixation, effectiveness in soil and water conservation, and high palatability, making it well-suited for both forage and ecological grass applications. The ecological significance of R. ciliaris is particularly exemplified by the elite cultivar ‘Liao sheng’, which was selected and developed by our research group. Characterized by its rapid spring growth and vigorous root system, ‘Liao sheng’ has been officially certified and extensively deployed for wind-break and sand-fixation in the semi-arid sandy regions of Northwest Liaoning. These successful applications underscore its critical role in soil and water conservation. Furthermore, R. ciliaris is a tetraploid wild relative of wheat, possessing extremely rich genetic diversity. It has evolved many elite characteristics that are either absent or lost in cultivated wheat over time through natural selection [1]. Currently, research on R. ciliaris mostly focuses on the germplasm resource collection and evaluation, genetic diversity, morphological anatomy, and breeding of disease-resistant crops. However, there are limited studies on its superior gene mining and functional characterization through omics and molecular biology techniques. Further research on the understanding of the molecular mechanisms of strong stress tolerance behind R. ciliaris will provide more valuable genetic resources for the genetic improvement of wheat and other important forage crops.
Investigating the expression patterns of genes is a prerequisite for in-depth functional genomics research [2]. Quantitative reverse transcription polymerase chain reaction (qRT-PCR) is a widely used technique for analyzing relative mRNA expression levels, utilizing complementary DNA (cDNA) as its template [3,4]. The reliability of this technique depends on several critical factors: high-quality cDNA template, primer specificity, high amplification efficiency, and appropriate selection of reference genes (RGs) [5]. To ensure accurate qRT-PCR data, normalization of target gene expression is essential. Currently, the most common normalization approach employs stable RGs as internal controls to minimize experimental variation [6]. Consequently, the accuracy of the selected RGs fundamentally determines the accuracy of the overall qRT-PCR analysis.
Numerous studies have identified RGs for qRT-PCR across diverse plant species, tissues, developmental stages, and environmental conditions [7]. However, research indicates that the expression levels of many commonly used RGs exhibit instability when compared across different experimental tissues or developmental stages [8]. For example, in Kobresia littledalei, ACTIN was the most stable RG in leaves and stems under control conditions, while GAPDH was optimal in roots [9]. In addition, in barnyard millet (Echinochloa crus-galli), the most stable RGs varied significantly with stress type: UBC5 and α-TUB under drought, GAPDH under salinity, GAPDH and APRT under cold, and EF-1α and RPII under heat stress [10]. These examples demonstrate that no single reference gene is universally stable across all plant species or experimental conditions. Therefore, to address this gap in R. ciliaris, our study hypothesized that stable reference genes could be identified and validated through a systematic evaluation, aiming to establish a reliable normalization method for gene expression studies in this species.
In this study, eight candidate RGs, namely Actin depolymerizing factor 4 like (ADF4L), Tubulin alpha (TUBA), Ubiquitin-like protein 5 (UBL5), Ubiquitin-conjugating enzyme (UCE2), Elongation factor 1 alpha (EF1A), Malate dehydrogenase (MDH), Ribosomal protein L19 (RPL19) and Glyceraldehyde-3-phosphate dehydrogenase (GAPDH), that are extensively used for qRT-PCR normalization in monocots were identified from the transcriptome data of R. ciliaris in response to drought stress. The expression profiles of these genes were quantified by qRT-PCR for different tissues (Leaves, roots, internode 1–3, young spikes, and mature spikes) and under drought stress conditions. The expression stability of candidate RGs was further evaluated using the ΔCt [11], GeNorm [12], NormFinder [13], BestKeeper [14], and RefFinder [15] algorithms. Finally, the gene encoding pyrroline-5-carboxylate synthase (P5CS) and protein phosphatase 2C (PP2C), which are key enzymes involved in proline biosynthesis and stress signaling pathways in plants, respectively, were used to confirm the expression stability of the most stable and least stable RGs in different tissues and under drought stress by qRT-PCR analysis. Our study systematically identifies the most stable RGs for R. ciliaris across various tissues and under drought stress, facilitating further research on the molecular mechanisms of stress resistance and functional analysis in R. ciliaris.

2. Materials and Methods

2.1. Plant Materials and Stress Treatments

Seeds of R. ciliaris accession ‘Liao sheng’ were collected from the National Ecological Grass Germplasm Resource Garden (Shenyang) at Shenyang Agricultural University, Liaoning Province, Northeast China. The seeds were surface-sterilized in a 5% sodium hypochlorite solution for 3 h, rinsed 4–5 times with sterile water, and placed in Petri dishes lined with filter paper moistened with 5 mL of sterile water. After 14 days of germination, seedlings were transferred to a hydroponic system containing half-strength Hoagland’s nutrient solution, which was renewed every 7 days. Seedlings were grown in the growth chamber at 24 ± 2 °C with 16 h light (390 μmol·m−2·s−1) for three weeks. Subsequently, seedlings with uniform growth performance were subjected to drought stress treatment by supplying 20% PEG 6000 solution, which generates an osmotic potential of approximately −0.59 MPa to simulate moderate osmotic stress [16]. Leaves and roots were harvested at 0, 4, 24, and 48 h post-treatment, flash-frozen in liquid N2, and stored at −80 °C. The samples were later sent for transcriptome sequencing and used for qRT-PCR.
For gene expression analysis across various tissues of R. ciliaris ‘Liao sheng’, samples were collected from both controlled indoor and field environments. Specifically, leaves and roots were obtained from 21-day-old hydroponically grown seedlings. The first, second, and third internodes (designated as internode 1, internode 2, and internode 3, respectively) and young spikes were collected from plants grown in the nursery of Shenyang Agricultural University at the booting stage, while mature spikes were collected at the heading stage, according to the growth stage criteria established by Moore et al. [17] (Figure S1). For both experimental setups described above, the biological replication was designed as follows: for the drought stress treatment, a sample from each time point was constituted as a pooled mixture from eight individual plants, representing one biological replicate; for the multi-tissue analysis, a sample of each tissue type was similarly constituted as a pooled mixture from eight individual plants, forming one biological replicate. Three such independent biological replicates were established for each experimental condition.

2.2. Reference Gene Selection and Primer Design

Candidate RGs were selected based on RNA-seq data obtained from R. ciliaris leaf and root samples under drought stress. Specifically, the standard deviation (SD) and coefficient of variation (CV) of gene expression levels were calculated from average FPKM values derived from leaf and root samples across all time points. Genes were filtered using the following stringent criteria: FPKM ≥ 50, CV < 0.5, |log2(fold change)| < 2, τ < 0.5 (Table S1), and prior documentation as reliable RGs in other monocot species [18,19]. Based on these criteria, eight candidate RGs (ADF4L, TUBA, UBL5, UCE2, EF1A, MDH, RPL19, and GAPDH) were selected for subsequent qRT-PCR validation in R. ciliaris.
Gene-specific primers (Table 1) were designed using Primer3Plus [20] and synthesized by Sangon Biotech Co., Ltd. (Shanghai, China). To evaluate amplification performance, mixed cDNAs were serially diluted in a five-fold gradient (1:5, 1:25, 1:125, 1:625, and 1:3125) to generate standard curves. The amplification efficiency (E) (E = (10[−1/slope] − 1) × 100) and correlation coefficient (R2) for each primer pair were automatically calculated using Bio-Rad CFX Manager™ Software v3.0. Primer specificity was confirmed by both RT-PCR (Figure S2) and melting curve analysis (Figure 1).

2.3. RNA Extraction, cDNA Synthesis, and qRT-PCR

Total RNA was extracted from drought-stressed and various tissue samples using the cetyltrimethylammonium bromide (CTAB) method. RNA concentration and purity were measured with a NanoDrop 2000 spectrophotometer, with all samples yielding A260/A280 ratios between 1.8 and 2.1, confirming high purity for downstream applications. RNA integrity was further verified by 1% agarose gel electrophoresis (Figure S2). First-strand cDNA was synthesized from 1 µg of total RNA using the PrimeScript™ RT reagent Kit with gDNA Eraser (TaKaRa, Dalian, China) according to the manufacturer’s instructions. All cDNA samples were stored at −20 °C for subsequent analysis.
Quantitative real-time PCR (qRT-PCR) was conducted using a CFX Duet Real-Time PCR System (Bio-Rad, Hercules, CA, USA). Reactions were performed in a 10 μL volume containing 5 μL of TB Green Premix EX Taq™ II (TaKaRa, Dalian, China), 1 μL of each forward and reverse primer (2 μM), 2 μL of 1:4 diluted cDNA template, and 1 μL of nuclease-free water. The thermal cycling protocol consisted of an initial denaturation at 95 °C for 30 s, followed by 40 cycles of 95 °C for 5 s and 60 °C for 30 s. Melting curve analysis was performed by heating from 65 °C to 95 °C in 0.5 °C increments. The analysis confirmed a single peak for each primer pair, indicating specific amplification without primer-dimers or non-specific products. Each sample was run in two technical replicates with three independent biological replicates.

2.4. Evaluating the Expression Stability of Candidate Reference Genes

The expression stability of the eight candidate RGs across all R. ciliaris samples was evaluated using four commonly used algorithms (ΔCt [11], GeNorm [12], NormFinder [13], and BestKeeper [14]) based on raw Ct values. The results from these methods were subsequently integrated and comprehensively ranked using RefFinder [15] to determine the most suitable RGs for qRT-PCR studies in R. ciliaris. The optimal number of RGs for qPCR normalization was determined by geNorm analysis. All statistical analyses using the aforementioned packages were performed according to the manufacturers’ protocols.

2.5. Validation of the Candidate Reference Genes

To validate the stability of the selected RGs, P5CS (a key gene encoding a rate-limiting enzyme in proline biosynthesis) and PP2C (a core component of the ABA signaling pathway) were chosen as the target genes based on transcriptome data. Their expression pattern were analyzed using both the most stable RGs (MDH and RPL19) and the least stable ones (GAPDH and TUBA) for comparison. The qRT-PCR primers for P5CS were as follows: forward, 5′-GACTCGCTTCGGTCTAGGTG-3′; reverse, 5′-TACACAACTCCCTTGTCGCC-3′. The primers for PP2C were as follows: forward, 5′-GACGAAGAACAGCTCGGACA-3′; reverse, 5′-CCAGCTGTTATTTCCACAACCC-3′. The qRT-PCR assay was performed as detailed in Section 2.3. Each sample was analyzed with three biological and two technical replicates. Transcriptomic FPKM values for P5CS and PP2Cserved as independent references for expression trends. Relative expression levels of P5CS and PP2C were calculated using the 2−ΔΔCt method [21].

2.6. Statistical Analysis

All statistical analyses were performed with IBM SPSS 26.0 software (SPSS Inc., Chicago, IL, USA). For multiple comparisons, one-way ANOVA followed by Tukey’s post hoc test was used, with a significance level set at p < 0.05.

3. Results

3.1. Primer Specificity and PCR Amplification Efficiency

Eight candidate RGs (ADF4L, TUBA, UBL5, UCE2, EF1A, MDH, RPL19, GAPDH) and two target genes (P5CS and PP2C) of R. ciliaris were selected for qRT-PCR normalization. The primer sequences for the candidate RGs were listed in Table 1. Reverse transcription polymerase chain reaction (RT-PCR) was first carried out to detect the specificity of each primer pair. The results showed that all RGs yielded a single amplicon, with sizes ranging from 100 to 250 bp (Figure S2). Melting curve analysis characterizes the dissociation behavior of double-stranded DNA during thermal denaturation [22]. The specificity of each RG primer was further confirmed by melting curve analysis using qRT-PCR. As shown in Figure 1, the melting curve of all primer pairs displayed a single amplification peak. The amplification efficiencies of all these candidate RGs ranged from 103.98% (TUBA) to 110.84% (UCE2), and the correlation coefficients varied from 0.9908 (UCE2 and RPL19) to 0.9970 (TUBA) (Figure S3 and Table 1).

3.2. Threshold Cycle (Ct) Values of Candidate RGs

The expression stability of the eight candidate RGs was evaluated by qRT-PCR across all experimental samples based on mean Ct values. Analysis was performed on three sample sets: (1) a tissue set comprising leaf, root, internodes 1–3, young spike, and mature spike under control conditions; (2) a drought stress set including leaf and root samples under drought treatment; and (3) a combined set containing all aforementioned samples. Representative images of the various R. ciliaris tissues are provided in Figure S1. The same grouping strategy was consistently applied for all subsequent stability evaluations, including analyses with the ΔCt, GeNorm, NormFinder, BestKeeper, and RefFinder algorithms.
The mean Ct values of these genes ranged from 16.83 (EF1A) to 26.24 (ADF4L), with the majority concentrated between 18 and 22 across all samples (Figure 2). Among all candidates, MDH exhibited the highest expression stability under all tested conditions, showing a consistent mean Ct value of 19.68 ± 0.10 with the smallest standard error (SE) (Figure 2; Table S3). In contrast, TUBA was the least stable gene across tissue types (mean Ct = 20.91 ± 0.33) and in the combined sample set (mean Ct = 19.93 ± 0.28), while GAPDH (mean Ct = 17.91 ± 0.37) showed the lowest stability under drought stress (Figure 2; Table S3). These results demonstrate that MDH exhibited minimal expression variation across all experimental conditions, while TUBA and GAPDH showed the most substantial fluctuations in all samples and under drought stress, respectively. Notably, GAPDH expression was undetectable in various tissues other than the leaf and root; therefore, its Ct value is only presented in the drought stress dataset.

3.3. Expression Stability of Candidate RGs

3.3.1. ΔCt Analysis

The ΔCt method evaluates expression stability by calculating the pairwise variation in Ct values between each gene and all other candidates across all samples, followed by determination of the standard deviation (SD) for each gene. A lower SD value indicates higher expression stability [23]. In the present study, MDH showed the lowest SD values across all samples and in the tissue set, confirming it as the most stable candidate under these conditions. Under drought stress, RPL19 displayed optimal stability. In contrast, GAPDH, UBL5, and TUBA were identified as the least stable genes under drought stress, across all samples, and among different tissues, respectively (Table 2).

3.3.2. GeNorm Analysis

The GeNorm algorithm assesses RG stability based on an M-value threshold (M < 1.5), with lower M-values indicating higher expression stability and better ranking [19]. Our results showed that all candidate RGs exhibited M-values below 1.5 under all experimental conditions. Among them, MDH and RPL19 (M = 0.37) were identified as the most stable genes across all sample groups. In contrast, UBL5 (M = 1.31) and GAPDH (M = 1.31) showed the lowest stability in different tissues and under drought stress, respectively, while TUBA (M = 1.42) was the least stable gene overall (Table 2). To quantitatively validate the reliability of these findings, Spearman’s rank correlation analysis was performed across the five algorithms (ΔCt, geNorm, NormFinder, BestKeeper, and RefFinder). The results demonstrated a high degree of consensus (Figure S4). Notably, geNorm exhibited near-perfect consistency with NormFinder, ΔCt, and RefFinder, with correlation coefficients (r) ranging from 0.98 to 1.00. Although its correlation with BestKeeper was slightly lower (r = 0.86) due to the latter’s reliance on raw variance (SD and CV), the overall agreement remained robust and statistically significant (p < 0.01). These results consistently demonstrate that MDH and RPL19 represent the most suitable reference genes for normalization in R. ciliaris.
The accuracy of gene expression quantification often requires the use of multiple RGs, as reliance on a single RG may lead to normalization bias [24]. To determine the optimal number of RGs, we applied the GeNorm algorithm, which calculates pairwise variation (Vn/(n + 1)) with a recommended threshold of 0.15 [25]. A value above this threshold indicates the need for (n + 1) RGs to minimize normalization error, while a value below it suggests that n RGs are sufficient [25]. In this study, all sample sets showed Vn/(n + 1) values exceeding 0.15 (Figure 3). Nevertheless, the 0.15 threshold should be interpreted as a guideline rather than a strict cutoff [21]. In cases where the Vn/(n + 1) remains above 0.15 or where reducing it below this value would require an impractical number of RGs, the use of the two or three most stable RGs represents an effective and practical normalization strategy [26,27]. In this study, we compared the normalization efficiency of the two most stable RGs (MDH and RPL19) against a combination of three RGs (MDH, RPL19, and EF1A). As shown in the Supplementary Figure S6, the relative expression patterns of RcP5CS obtained using the two-RG combination were highly consistent with those obtained using the three-RG combination, with no significant differences observed in the statistical grading. These results demonstrate that the combination of MDH and RPL19 is sufficient to provide accurate and reliable normalization. Consequently, although the GeNorm V values (such as V3/4 of 0.18 in tissues and 0.16 under drought stress) remained slightly above 0.15, the combination of MDH and RPL19 was selected as the optimal normalization strategy for subsequent qRT-PCR assays.

3.3.3. NormFinder Analysis

To further evaluate the stability of candidate RGs, we employed the NormFinder algorithm, which estimates expression stability by measuring both intra- and inter-group variations. Lower stability values (M) indicate higher expression stability [25]. In the combined sample set, RPL19 (M = 0.08) was the most stable RG, followed by MDH, EF1A, UCE2, ADF4L, and UBL5, while TUBA (M = 1.65) exhibited the highest variation. In both the tissue and drought stress groups, MDH demonstrated the greatest stability (M = 0.07 and M = 0.26, respectively), with RPL19 ranking second (M = 0.44 and M = 0.28, respectively). The least stable genes differed between groups: UBL5 (M = 1.53) was the most unstable in the tissue group, whereas GAPDH (M = 1.50) showed the lowest stability under drought stress (Table 2). In summary, consistent with the results derived from the GeNorm algorithm, MDH and RPL19 consistently ranked as the most optimal RGs, while TUBA, UBL5, and GAPDH were the least stable ones.

3.3.4. BestKeeper Analysis

The BestKeeper algorithm evaluates RG stability by calculating the standard deviation (SD) and coefficient of variation (CV) of qRT-PCR Ct values, where lower values indicate higher stability and an SD > 1.0 generally signifies unstable expression [3]. As summarized in Table 3, ADF4L and TUBA showed SD values exceeding 1.0 across all three sample sets, reflecting their consistently low expression stability (Table 3). Similarly, EF1A and GAPDH were ranked as the least stable genes in the combined sample set and under drought stress, respectively. In contrast, UBL5 was the most stable under drought conditions, while MDH demonstrated the highest stability across tissue types and in the combined sample set (Table 3).

3.3.5. RefFinder Analysis

The RefFinder platform integrates results from the ΔCt, GeNorm, NormFinder, and BestKeeper algorithms to generate a comprehensive stability ranking of candidate RGs [28]. According to the RefFinder analysis, MDH was identified as the most stable RG across all experimental groups, which aligns with its consistent top ranking in the tissue sample set by all four individual algorithms (Figure 4). The overall stability order of the candidate genes across all samples was: MDH > RPL19 > EF1A > UCE2 > ADF4L > UBL5 > TUBA. Furthermore, GAPDH and TUBA were determined to be the least stable RGs under drought stress and across different tissues, respectively (Figure 4 and Table 4).

3.4. Validation of the Candidate RGs

As a key enzyme in proline biosynthesis, Pyrroline-5-carboxylate synthase (P5CS) has been consistently used as a molecular marker for monitoring drought and salinity stress in plants [29,30]. To validate the performance of candidate RGs, we analyzed the expression of P5CS as a target gene via qRT-PCR under tissue-specific and drought stress conditions. The two most stable RGs (MDH and RPL19), the two least stable RGs (GAPDH and TUBA), and the combination of the top two RGs were selected to normalize P5CS expression (Figure 5). When normalized to the stable single RGs (MDH or RPL19) or their combination, the expression profile of P5CS closely aligned with transcriptome data, showing a marked up-regulation in leaves under drought stress that peaked at 24 h after treatment initiation (Figure 5A,B). To quantitatively validate this agreement, we performed a linear regression analysis between the transcript abundance (FPKM) and the qRT-PCR relative expression levels. The results demonstrated a high degree of consistency, with a determination coefficient (R2) of 0.8948 for the optimal RG combination in leaves (Figure S5A). To rigorously evaluate the necessity of including additional reference genes, we compared the normalization factor derived from the top two RGs with that derived from the top three RGs (including EF1A). The overall expression patterns and statistical significance levels of P5CS relative expression normalized by these two combinations remained highly consistent (Figure S6). Given that the inclusion of EF1A did not alter the biological interpretation of the stress response, the dual-gene combination of MDH and RPL19 was selected as the optimal normalization strategy to balance accuracy and experimental efficiency. In roots, transcriptome data indicated a slight but non-significant induction of P5CS after 4 h of drought, a trend consistently reproduced using MDH or RPL19 (individually or combined) for normalization (Figure 5C). This consistency was further statistically corroborated by linear regression analysis, which yielded an impressive determination coefficient (R2) of 0.9668 for roots (Figure S5B). In contrast, normalization with the least stable gene, GAPDH, failed to detect statistically significant changes in P5CS expression in leaves and produced expression trends inconsistent with transcriptome data in roots (Figure 5B,C).
In addition to its expression under drought stress, the transcript profile of P5CS was also analyzed across various tissues of R. ciliaris under normal growth conditions. When normalized using the optimal RG combination (MDH + RPL19), P5CS expression was predominantly observed in leaves, showing a significant 10.3-fold increase compared to roots (p < 0.05), followed by mature spikes (Figure 5D). In contrast, normalization with the least stable gene, TUBA, yielded an erratic expression profile; although it indicated a nominally higher difference (21.5-fold), this variation failed to reach statistical significance (p > 0.05) due to the high instability of TUBA, thereby obscuring the true tissue-specificity. Therefore, MDH and RPL19 are established as reliable RGs for accurate gene expression normalization in R. ciliaris across tissues and under drought stress conditions.
To further substantiate the reliability of the selected reference genes, the expression profile of Protein Phosphatase 2C (PP2C), a critical negative regulator of ABA signaling, was investigated under drought stress and across various tissues. Consistent with the transcriptome data (FPKM), which showed a rapid and sustained induction of PP2C in leaves and a gradual increase in roots under drought conditions (Figure 6A), normalization using the two most stable RGs (MDH and RPL19) and their combination accurately reproduced these expression patterns (Figure 6B,C). Linear regression analysis quantitatively confirmed this consistency, revealing distinct positive correlations between RNA-seq FPKM values and relative expression levels in both drought-stressed leaves (R2 = 0.7624, Figure S5C) and roots (R2 = 0.8871, Figure S5D). Specifically, in leaves, PP2C expression normalized by the optimal RGs exhibited a significant up-regulation starting at 4 h and peaking at 24 h. In contrast, normalization with the least stable gene, GAPDH, resulted in a notable underestimation of PP2C abundance, particularly in roots at 4 h and 24 h, where it failed to capture the significant induction observed with the stable RGs (Figure 6C).
In addition to stress responses, the tissue-specific expression of PP2C was analyzed to assess normalization accuracy under normal growth conditions. When normalized using the optimal combination of MDH and RPL19, PP2C exhibited high expression levels in internodes (especially Internodes 1 and 3) and leaves, with moderate levels in spikes (Figure 6D). However, the use of the least stable gene, TUBA, generated a contradictory expression profile. Most notably, TUBA-based normalization failed to detect the high abundance of PP2C in the internodes, falsely presenting them as low-expression tissues, and disproportionately emphasized expression in the leaves. These discrepancies further confirm that the use of unstable reference genes can lead to erroneous biological interpretations, reinforcing the suitability of MDH and RPL19 as the optimal normalization strategy for R. ciliaris.

4. Discussion

R. ciliaris is a drought-tolerant grass species native to China. Furthermore, this species exhibits multiple advantageous traits, such as tolerance to barren and cold conditions and vigorous seedling growth in spring, making it well-suited for pasture improvement and ecological restoration [31]. Our research group developed ‘Liao sheng’, an elite cultivar selected from wild R. ciliaris accessions and well-suited for ecological restoration in semi-arid sandy lands of western Liaoning. However, the molecular mechanisms underlying its drought tolerance remain poorly characterized. Investigating these mechanisms is further complicated by the fact that R. ciliaris is an allotetraploid species (2n = 4x = 28, genome ScScYcYc) [1]. The potential for homeolog expression bias and gene dosage effects inherent to polyploids necessitates a rigorously validated reference gene system to accurately distinguish true biological stress responses from genomic noise [32,33]. As a highly accurate and sensitive technique, qRT-PCR is indispensable for quantifying gene expression and deciphering regulatory networks [34]. Its reliable application in R. ciliaris, however, depends on the use of stably expressed RGs as well as stringent primer specificity. Ideal RGs should exhibit consistent expression across tissues and experimental conditions, unaffected by external stimuli or internal factors [35]. To date, systematic RG selection has not yet been established for R. ciliaris. To address this gap, we selected eight candidate RGs from transcriptome data and evaluated their stability using multiple algorithms to identify optimal RGs for use across different tissues and under drought stress.
The integration of multiple analytical approaches is essential for obtaining a robust evaluation of RG stability [36]. In this study, we systematically evaluated candidate RGs using five distinct algorithms (ΔCt, geNorm, NormFinder, BestKeeper, and RefFinder). The results revealed a broad consensus with minor algorithm-dependent variations. Specifically, all algorithms consistently identified TUBA as the least stable gene, while MDH was ranked as the most stable by four of the five methods (all except NormFinder) under our experimental conditions. However, a slight divergence was observed with NormFinder, which ranked MDH as the second most stable gene. This discrepancy is likely attributed to the distinct statistical principles of the algorithms, as NormFinder accounts for both intra- and inter-group variations, while geNorm relies on pairwise stability [25]. Despite this, the comprehensive ranking by RefFinder confirmed MDH as the overall most stable RG. This finding is consistent with previous work in Macadamia integrifolia, where MDH was also identified as stable RG across multiple tissues using the same five algorithms [12]. Interestingly, a contrasting result was reported in apple, where MDH was ranked as the least stable gene across all four algorithms applied (geNorm, BestKeeper, NormFinder, and RefFinder) [37]. This divergence highlights that RG stability is not an intrinsic property of the gene itself but is contingent upon the specific transcriptional networks and metabolic adaptations of the host organism. Since MDH is a key enzyme in the TCA cycle, its expression patterns can vary significantly depending on how different species regulate their energy metabolism under specific physiological constraints [38]. This divergence suggests that RG stability is highly context-dependent. Therefore, employing a multi-algorithm approach is essential to obtain reliable and convergent results for specific experimental conditions [39].
In this study, MDH and RPL19 were identified as the two most stable RGs. Malate dehydrogenase (MDH), which catalyzes the reversible conversion of malate to oxaloacetate, is widely distributed across multiple cellular compartments in plants [40]. While MDH demonstrated the highest expression stability across all experimental groups in R. ciliaris, its performance varies among species. Consistent with our findings, MDH has been reported as a stable RG in the aerial and tuber parts of tiger nut, as well as under MeJA treatment conditions in M. integrifolia [12,24]. In contrast, it showed lower stability in Baphicacanthus cusia across different organs and under MeJA treatment [41], as well as in the peel and flesh of apple [37], indicating potential species-specific regulatory mechanisms. The ribosomal protein L19 (RPL19), a core component of the eukaryotic translation machinery [42], was identified as the second most stable RG in this study. This finding aligns with its documented consistent stability across various tissues in kiwifruit and potato [43,44], supporting its broad utility as a reliable normalization gene in plant species. Although glyceraldehyde-3-phosphate dehydrogenase (GAPDH) is widely employed as a RG in various plant species due to its central role in carbohydrate metabolism [45], our study revealed markedly unstable expression of GAPDH under drought stress in R. ciliaris. This instability is likely attributable to the multifaceted role of GAPDH in plant stress adaptation. Beyond its canonical function in glycolysis, GAPDH acts as a ‘moonlighting’ protein that is actively regulated to modulate energy homeostasis and cellular signaling in response to abiotic stimuli [46]. Consequently, the physiological demand for metabolic reconfiguration during drought adaptation compromises its suitability as a stable reference. This finding is consistent with reported instability of GAPDH expression in other plant species subjected to abiotic stress, including annual ryegrass and cannabis [47,48]. TUBA (α-tubulin) is crucial for microtubule formation in eukaryotic cells and plays a key role in maintaining cellular structure and function [49]. It has been identified as one of the most stable RGs in species such as Schisandra chinensis and oat [50,51]. By contrast, in the present study, TUBA exhibited the lowest expression stability across tissues and all sample sets in R. ciliaris. This result aligns with previous reports of its instability in Rosa praelucens, M. integrifolia, and Angelica decursiva [52,53,54], further highlighting the species-specific nature of RG performance.
Actin and Ubiquitin have been extensively utilized as RGs in gene expression studies across monocotyledonous Poaceae species under diverse abiotic stresses and tissue types [18,55,56]. Actin serves as a core structural component of microfilaments in plant cells and is essential for critical physiological functions such as cell shape maintenance and cytoplasmic streaming [57]. Ubiquitin defines a structurally conserved protein family that plays a central regulatory role in essential eukaryotic processes including protein degradation, signal transduction, and stress responses [58]. Previous studies have demonstrated the high expression stability of both Actin and Ubiquitin in drought-stressed wheat [19], with similar stability patterns reported in other crops such as rice and maize under various stress conditions [59,60]. Furthermore, while the wheat Actin ortholog has been used as an internal control for studying RcDREB1 expression in R. ciliaris under abiotic stress [61], our transcriptomic data identified only the ACTIN3, whose FPKM values, however, did not meet the established screening criteria. Therefore, ACTIN3 was excluded as a candidate RG in this study. It is noteworthy that the dynamic reorganization of the actin cytoskeleton is regulated by Actin Depolymerizing Factor (ADF), which facilitates actin network remodeling by severing filamentous actin (F-actin) and promoting the dissociation of globular actin (G-actin) monomers [62]. Correspondingly, ADF4-like (ADF4L) in our study was found to be unstable across all three experimental groups. While ubiquitin-like protein 5 (UBL5) shares structural similarity with ubiquitin, it functions independently of the canonical ubiquitination pathway, instead regulating target proteins through isopeptide bond formation [63]. In the present study, UBL5 exhibited low expression stability across all three experimental groups. This result contrasts with its reported stability in tiger nut (Cyperus esculentus) and Saussurea laniceps [24,50], but aligns with its unstable performance in seashore paspalum [64]. These divergent findings underscore the critical importance of RG selection based on experimental validation under specific species, tissue types, and experimental conditions, rather than relying on common practice or isolated reports [65].
GeNorm determines the optimal pair of RGs by calculating the pairwise variation (Vn/(n + 1)), with the default threshold of V < 0.15 indicating that no additional RGs are required [25]. In this study, all V values exceeded the 0.15 threshold. This outcome is not uncommon in RG screening in plants, as demonstrated in both Pyropia yezoensis [14] and grapevine [66], where a pair of RGs was successfully adopted as the optimal combination despite V > 0.15. However, strictly adhering to the 0.15 threshold is often viewed as pragmatic rather than obligatory. To rigorously justify the sufficiency of two RGs, we validated our selection by comparing the normalization factors derived from the top two RGs (MDH and RPL19) with those from the top three RGs (including EF1A). The analysis of the target gene P5CS demonstrated that the inclusion of the third gene did not result in any statistically significant differences in expression patterns or levels (Figure S6). This confirms that adding a third gene contributes no additional analytical precision but merely increases experimental complexity. Therefore, based on a comprehensive evaluation of stability, reproducibility, and statistical validation, MDH and RPL19 were ultimately selected as the optimal RG pair for this study.
To confirm the reliability of the RGs selected by the five algorithms, we analyzed the relative expression of the target genes (P5CS and PP2C) under drought stress and across various tissues in R. ciliaris, using both the most stable (MDH and RPL19) and least stable (GAPDH and TUBA) genes for normalization. qRT-PCR results showed that the expression of both P5CS and PP2C was significantly induced by drought in leaves when normalized with MDH or RPL19, either individually or in combination, closely matching the transcriptomic data. Furthermore, the combined use of MDH and RPL19 provided statistically more robust results. In contrast, normalization with GAPDH failed to accurately capture the peak expression levels at 24h or the specific induction patterns of these genes under stress. Tissue-specific analysis confirmed that the use of MDH and RPL19 (individually or combined) consistently detected distinct expression patterns for the two target genes: significantly higher expression of P5CS was observed in leaves and mature spikes, whereas PP2C exhibited higher expression levels in internode 1 and mature spikes. In both cases, the lowest expression was consistently detected in roots. These findings align with the expression patterns reported in switchgrass [29] and reed canary grass [67]. Conversely, normalization with TUBA led to inconsistent statistical results and failed to identify such tissue-specific differences.

5. Conclusions

This study represents the first systematic identification and validation of RGs in R. ciliaris, with comprehensive analyses of P5CS and PP2C expression patterns across various tissues and under drought stress conditions. Beyond merely identifying suitable candidates, these findings resolve a critical methodological bottleneck that has historically limited the accuracy of transcriptomic quantification in this species. By providing a standardized normalization protocol, this study ensures the reliability of future investigations into complex stress-responsive pathways, such as the ABA signaling and proline biosynthesis networks highlighted by our P5CS and PP2C analyses. Ultimately, robust gene expression profiling is a prerequisite for uncovering novel functional genes, thereby accelerating molecular breeding programs aimed at enhancing drought resilience and ecological restoration capabilities of forage grasses in arid environments.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/genes17020237/s1, Figure S1: Phenotype diagrams of different tissues of R. ciliaris ‘Liao sheng’; Figure S2: Analysis of PCR amplification specificity for the eight candidate reference genes via agarose gel electrophoresis; Figure S3: Standard curves of the eight candidate reference genes; Figure S4: Spearman’s rank correlation heatmap of candidate reference genes ranking; Figure S5: Correlation analysis validating the consistency between RNA-seq (FPKM) and qRT-PCR expression profiles for RcP5CS and RcPP2C under drought stress; Figure S6: Comparison of RcP5CS relative expression normalized by the top two and top three reference genes; Table S1: FPKM values of the eight reference genes in leaf and root samples under drought stress; Table S2: Log2 fold-change values of the eight reference genes under drought stress; Table S3: The Ct values of the eight reference genes in R. ciliaris across various tissues and under drought stress.

Author Contributions

S.L. and Q.L. designed research; Q.L., Y.L. (Yue Liu), Y.W., G.Z., J.L., H.L. and Z.L. performed research; Q.L. and Y.L. (Yue Liu) analyzed data; Y.L. (Ying Liu) and L.B. provided seeds of R. ciliaris ‘Liao sheng’; Y.L. (Ying Liu), L.B. and S.L. supervised research; and Q.L. and S.L. wrote the paper. All authors have read and agreed to the published version of the manuscript.

Funding

This work and the APC were funded by the Educational Commission of Liaoning Province of China (LJKMZ20221055) and the Talent Introduction Project of Shenyang Agricultural University (X2022004).

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Dataset available on request from the authors: The raw data supporting the conclusions of this article will be made available by the authors on request.

Acknowledgments

We thank all authors of this study.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Song, R.; Cheng, Y.; Wen, M.; Song, X.; Wang, T.; Xia, M.; Sun, H.; Cheng, M.; Cui, H.; Yuan, C.; et al. Transferring a new Fusarium head blight resistance locus FhbRc1 from Roegneria ciliaris into wheat by developing alien translocation lines. Theor. Appl. Genet. 2023, 136, 36. [Google Scholar] [CrossRef]
  2. Wei, H.S.; Qiao, H.L.; Liu, S.; Yuan, X.Q.; Xu, C.Q. Transcriptome-Based Selection and Validation of Reference Genes for Gene Expression in Goji Fruit Fly (Neoceratitis asiatica Becker) Under Developmental Stages and Five Abiotic Stresses. Int. J. Mol. Sci. 2023, 24, 451. [Google Scholar] [CrossRef]
  3. Pfaffl, M.W.; Tichopad, A.; Prgomet, C.; Neuvians, T.P. Determination of stable housekeeping genes, differentially regulated target genes and sample integrity: BestKeeper-Excel-based tool using pair-wise correlations. Biotechnol. Lett. 2004, 26, 509–515. [Google Scholar] [CrossRef]
  4. Perez-Arellano, I.; Carmona-Alvarez, F.; Martinez, A.I.; Rodriguez-Diaz, J.; Cervera, J. Pyrroline-5-carboxylate synthase and proline biosynthesis: From Osmotolerance to rare metabolic disease. Protein Sci. 2010, 19, 372–382. [Google Scholar] [CrossRef]
  5. Yang, Z.; Zhao, X.; Shang, W.; Liu, Y.; Ji, J.-F.; Liu, J.-P.; Tong, C. Pyrroline-5-carboxylate synthase senses cellular stress and modulates metabolism by regulating mitochondrial respiration. Cell Death Differ. 2021, 28, 303–319. [Google Scholar] [CrossRef]
  6. Long, X.-Y.; Wang, J.-R.; Ouellet, T.; Rocheleau, H.; Wei, Y.-M.; Pu, Z.-E.; Jiang, Q.-T.; Lan, X.-J.; Zheng, Y.-L. Genome-wide identification and evaluation of novel internal control genes for Q-PCR based transcript normalization in wheat. Plant Mol. Biol. 2010, 74, 307–311. [Google Scholar] [CrossRef]
  7. Mascia, T.; Santovito, E.; Gallitelli, D.; Cillo, F. Evaluation of reference genes for quantitative reverse-transcription polymerase chain reaction normalization in infected tomato plants. Mol. Plant Pathol. 2010, 11, 805–816. [Google Scholar] [CrossRef]
  8. Zhang, L.; Zhang, R.; Hu, X.; Li, J.; Zhao, P.; Guo, F.; Zhao, H.; Wang, P.; Wang, Y.; Ni, D.; et al. Reference gene selection for qRT-PCR analysis in the shoots and roots of Camellia sinensis var. sinensis under nutritional stresses. Sci. Hortic. 2023, 320, 112237. [Google Scholar] [CrossRef]
  9. He, M.; Cui, S.; Yang, X.; Mu, G.; Chen, H.; Liu, L. Selection of suitable reference genes for abiotic stress-responsive gene expression studies in peanut by real-time quantitative PCR. Electron. J. Biotechnol. 2017, 28, 76–86. [Google Scholar] [CrossRef]
  10. Jin, Y.H.; Liu, F.; Huang, W.; Sun, Q.; Huang, X.Z. Identification of reliable reference genes for qRT-PCR in the ephemeral plant (Arabidopsis pumila) based on full-length transcriptome data. Sci. Rep. 2019, 9, 11. [Google Scholar] [CrossRef] [PubMed]
  11. Chi, H.; Bai, S.; Chen, C.; Bao, Y.; Qu, X.; Sun, S.; Pan, J.; Yang, X.; Hou, C.; Deng, Y. Identification of reference genes via real-time quantitative PCR for investigation of the transcriptomic basis of the squalene biosynthesis in different tissues on olives under drought stress. Plant Stress 2024, 14, 100575. [Google Scholar] [CrossRef]
  12. Yang, Q.; Yang, Z.; Zeng, H.; Zou, M.; Song, X.; Wan, J.; Wang, Z.; Chen, J.; Luo, L. Evaluation and Validation of Reliable Reference Genes for Quantitative Real-Time PCR Analysis of the Gene Expression in Macadamia integrifolia. Forests 2024, 15, 1966. [Google Scholar] [CrossRef]
  13. Andersen, C.L.; Jensen, J.L.; Orntoft, T.F. 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. [Google Scholar] [CrossRef]
  14. Kong, F.N.; Cao, M.; Sun, P.P.; Liu, W.X.; Mao, Y.X. Selection of reference genes for gene expression normalization in Pyropia yezoensis using quantitative real-time PCR. J. Appl. Phycol. 2015, 27, 1003–1010. [Google Scholar] [CrossRef]
  15. Zhang, Y.T.; Peng, X.R.; Liu, Y.; Li, Y.L.; Luo, Y.; Wang, X.R.; Tang, H.R. Evaluation of suitable reference genes for qRT-PCR normalization in strawberry (Fragaria × ananassa) under different experimental conditions. BMC Mol. Biol. 2018, 19, 10. [Google Scholar] [CrossRef] [PubMed]
  16. Michel, B.E.; Kaufmann, M.R. The osmotic potential of polyethylene glycol 6000. Plant Physiol. 1973, 51, 914–916. [Google Scholar] [CrossRef]
  17. Moore, K.; Moser, L.E.; Vogel, K.P.; Waller, S.S.; Johnson, B.; Pedersen, J.F. Describing and quantifying growth stages of perennial forage grasses. Agron. J. 1991, 83, 1073–1077. [Google Scholar] [CrossRef]
  18. Kiarash, J.G.; Wilde, H.D.; Amirmahani, F.; Moemeni, M.M.; Zaboli, M.; Nazari, M.; Moosavi, S.S.; Jamalvandi, M. Selection and validation of reference genes for normalization of qRT-PCR gene expression in wheat (Triticum durum L.) under drought and salt stresses. J. Genet. 2018, 97, 1433–1444. [Google Scholar] [CrossRef] [PubMed]
  19. Dudziak, K.; Sozoniuk, M.; Szczerba, H.; Kuzdralinski, A.; Kowalczyk, K.; Börner, A.; Nowak, M. Identification of stable reference genes for qPCR studies in common wheat (Triticum aestivum L.) seedlings under short-term drought stress. Plant Methods 2020, 16, 8. [Google Scholar] [CrossRef]
  20. Untergasser, A.; Cutcutache, I.; Koressaar, T.; Ye, J.; Faircloth, B.C.; Remm, M.; Rozen, S.G. Primer3—New capabilities and interfaces. Nucleic Acids Res. 2012, 40, e115. [Google Scholar] [CrossRef]
  21. Livak, K.J.; Schmittgen, T.D. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 2001, 25, 402–408. [Google Scholar] [CrossRef]
  22. Wittwer, C.T. High-resolution DNA melting analysis: Advancements and limitations. Hum. Mutat. 2009, 30, 857–859. [Google Scholar] [CrossRef]
  23. Silver, N.; Best, S.; Jiang, J.; Thein, S.L. Selection of housekeeping genes for gene expression studies in human reticulocytes using real-time PCR. BMC Mol. Biol. 2006, 7, 33. [Google Scholar] [CrossRef]
  24. Bai, X.; Chen, T.; Wu, Y.; Tang, M.Y.; Xu, Z.F. Selection and Validation of Reference Genes for qRT-PCR Analysis in the Oil-Rich Tuber Crop Tiger Nut (Cyperus esculentus) Based on Transcriptome Data. Int. J. Mol. Sci. 2021, 22, 2569. [Google Scholar] [CrossRef]
  25. Vandesompele, J.; De Preter, K.; Pattyn, F.; Poppe, B.; Van Roy, N.; De Paepe, A.; Speleman, F. Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol. 2002, 3, RESEARCH0034.1. [Google Scholar] [CrossRef]
  26. De Ketelaere, A.; Goossens, K.; Peelman, L.; Burvenich, C. Technical note: Validation of internal control genes for gene expression analysis in bovine polymorphonuclear leukocytes. J. Dairy Sci. 2006, 89, 4066–4069. [Google Scholar] [CrossRef] [PubMed]
  27. Silveira, E.D.; Alves-Ferreira, M.; Guimaraes, L.A.; da Silva, F.R.; Carneiro, V.T.D. Selection of reference genes for quantitative real-time PCR expression studies in the apomictic and sexual grass Brachiaria brizantha. BMC Plant Biol. 2009, 9, 10. [Google Scholar] [CrossRef] [PubMed]
  28. Xie, F.L.; Wang, J.Y.; Zhang, B.H. RefFinder: A web-based tool for comprehensively analyzing and identifying reference genes. Funct. Integr. Genom. 2023, 23, 125. [Google Scholar] [CrossRef]
  29. Kishor, P.B.K.; Hong, Z.; Miao, G.H.; Hu, C.A.A.; Verma, D.P.S. Overexpression of delta-Pyrroline-5-Carboxylate Synthetase Increases Proline Production and Confers Osmotolerance in Transgenic Plants. Plant Physiol. 1995, 108, 1387–1394. [Google Scholar] [CrossRef] [PubMed]
  30. Szabados, L.; Savouré, A. Proline: A multifunctional amino acid. Trends Plant Sci. 2010, 15, 89–97. [Google Scholar] [CrossRef]
  31. Zhang, H.; Wu, X.; Guo, S.; Hong, T.; Xiong, J.; Lu, J.; Yang, L. Phenotypic diversity and comprehensive evaluation of Roegneria ciliaris (Trin.) Nevski germplasm resources in Henan. J. South. Agric. 2024, 55, 3003–3012. [Google Scholar]
  32. Grover, C.E.; Gallagher, J.P.; Szadkowski, E.P.; Yoo, M.J.; Flagel, L.E.; Wendel, J.F. Homoeolog expression bias and expression level dominance in allopolyploids. New Phytol. 2012, 196, 966–971. [Google Scholar] [CrossRef] [PubMed]
  33. Birchler, J.A.; Veitia, R.A. The gene balance hypothesis: From classical genetics to modern genomics. Plant Cell 2007, 19, 395–402. [Google Scholar] [CrossRef]
  34. Jozefczuk, J.; Adjaye, J. Quantitative real-time PCR-based analysis of gene expression. Methods Enzymol. 2011, 500, 99–109. [Google Scholar] [CrossRef] [PubMed]
  35. Tong, J.; Hu, M.; Han, B.; Ji, Y.; Wang, B.; Liang, H.; Liu, M.; Wu, Z.; Liu, N. Determination of reliable reference genes for gene expression studies in Chinese chive (Allium tuberosum) based on the transcriptome profiling. Sci. Rep. 2021, 11, 16558. [Google Scholar] [CrossRef]
  36. Lei, Y.; Yang, Z.; Zhou, Y.; Li, L. Selection of stable reference genes for qPCR-based analysis of circulating cell-free DNA levels in peripheral blood of neuroblastoma patients. Scand. J. Clin. Lab. Investig. 2025, 85, 590–598. [Google Scholar] [CrossRef] [PubMed]
  37. Zhu, L.F.; Yang, C.Q.; You, Y.H.; Liang, W.; Wang, N.N.; Ma, F.W.; Li, C.Y. Validation of reference genes for qRT-PCR analysis in peel and flesh of six apple cultivars (Malus domestica) at diverse stages of fruit development. Sci. Hortic. 2019, 244, 165–171. [Google Scholar] [CrossRef]
  38. Huang, Y.M.M.; Huber, G.A.; Wang, N.; Minteer, S.D.; McCammon, J.A. Brownian dynamic study of an enzyme metabolon in the TCA cycle: Substrate kinetics and channeling. Protein Sci. 2018, 27, 463–471. [Google Scholar] [CrossRef]
  39. Cheng, Y.; Pang, X.; Wan, H.J.; Ahammed, G.J.; Yu, J.H.; Yao, Z.P.; Ruan, M.Y.; Ye, Q.J.; Li, Z.M.; Wang, R.Q.; et al. Identification of Optimal Reference Genes for Normalization of qPCR Analysis During Pepper Fruit Development. Front. Plant Sci. 2017, 8, 14. [Google Scholar] [CrossRef]
  40. Wolyniak, M.J.; Frazier, R.H.; Gemborys, P.K.; Loehr, H.E. Malate dehydrogenase: A story of diverse evolutionary radiation. Essays Biochem. 2024, 68, 213–220. [Google Scholar] [CrossRef]
  41. Huang, Y.; Tan, H.; Yu, J.; Chen, Y.; Guo, Z.; Wang, G.; Zhang, Q.; Chen, J.; Zhang, L.; Diao, Y. Stable Internal Reference Genes for Normalizing Real-Time Quantitative PCR in Baphicacanthus cusia Under Hormonal Stimuli and UV Irradiation, and in Different Plant Organs. Front. Plant Sci. 2017, 8, 668. [Google Scholar] [CrossRef] [PubMed]
  42. Kang, J.; Brajanovski, N.; Chan, K.T.; Xuan, J.C.; Pearson, R.B.; Sanij, E. Ribosomal proteins and human diseases: Molecular mechanisms and targeted therapy. Signal Transduct. Target. Ther. 2021, 6, 22. [Google Scholar] [CrossRef]
  43. Chen, Y.L.; Yuan, Y.T.; Qiu, Z.H.; Hong, W.J.; Mao, K.; Yao, C.C.; Liu, Z.D.; Wang, N.N. Identification of reference genes for RT-qPCR analysis across kiwifruit species under iron deficiency conditions. Sci. Hortic. 2023, 312, 111865. [Google Scholar] [CrossRef]
  44. Li, G.; Zhou, Y.; Zhao, Y.Q.; Liu, Y.X.; Ke, Y.W.; Jin, X.Q.; Ma, H.L. Internal Reference Gene Selection for Quantitative Real-Time RT-PCR Normalization in Potato Tissues. Phyton-Int. J. Exp. Bot. 2020, 89, 328–344. [Google Scholar] [CrossRef]
  45. Lazarev, V.F.; Guzhova, I.V.; Margulis, B.A. Glyceraldehyde-3-phosphate Dehydrogenase Is a Multifaceted Therapeutic Target. Pharmaceutics 2020, 12, 416. [Google Scholar] [CrossRef]
  46. Cao, Y.P.; Hong, J.Y.; Wang, H.; Lin, M.F.; Cai, Y.P.; Liao, L.; Li, X.X.; Han, Y.P. Beyond glycolysis: Multifunctional roles of glyceraldehyde-3-phosphate dehydrogenases in plants. Hortic. Res. 2025, 12, uhaf070. [Google Scholar] [CrossRef]
  47. Liu, Q.X.; Qi, X.; Yan, H.D.; Huang, L.K.; Nie, G.; Zhang, X.Q. Reference Gene Selection for Quantitative Real-Time Reverse-Transcriptase PCR in Annual Ryegrass (Lolium multiflorum) Subjected to Various Abiotic Stresses. Molecules 2018, 23, 172. [Google Scholar] [CrossRef] [PubMed]
  48. Guo, R.; Guo, H.Y.; Zhang, Q.Y.; Guo, M.B.; Xu, Y.P.; Zeng, M.; Lv, P.; Chen, X.; Yang, M. Evaluation of reference genes for RT-qPCR analysis in wild and cultivated Cannabis. Biosci. Biotechnol. Biochem. 2018, 82, 1902–1910. [Google Scholar] [CrossRef] [PubMed]
  49. Roll-Mecak, A. The Tubulin Code in Microtubule Dynamics and Information Encoding. Dev. Cell 2020, 54, 7–20. [Google Scholar] [CrossRef]
  50. Liu, H.; Lu, Y.N.; Wang, X.J.; Wang, X.W.; Li, R.C.; Lu, C.F.; Lan, X.Z.; Chen, Y.Z. Selection and Validation of Reference Genes for RT-qPCR Analysis in Tibetan Medicinal Plant Saussurea Laniceps Callus Under Abiotic Stresses and Hormone Treatments. Genes 2022, 13, 904. [Google Scholar] [CrossRef]
  51. Jarosová, J.; Kundu, J.K. Validation of reference genes as internal control for studying viral infections in cereals by quantitative real-time RT-PCR. BMC Plant Biol. 2010, 10, 9. [Google Scholar] [CrossRef]
  52. Jian, H.Y.; Wang, H.C.; Qiu, X.Q.; Yan, H.J.; Ma, L.L. Identification and Validation of Reference Genes for qRT-PCR Analysis of Petal-Color-Related Genes in Rosa praelucens. Genes 2024, 15, 277. [Google Scholar] [CrossRef]
  53. Yang, Q.; Yang, Z.; Zhou, Y.; Chen, D.; Heng, L. Screening of stable reference genes for qRT-PCR analysis in Macadamia integrifolia. Chin. J. Trop. Crops 2020, 41, 1505–1512. [Google Scholar] [CrossRef]
  54. He, Y.D.; Zhong, Y.; Bao, Z.Z.; Wang, W.Q.; Xu, X.Q.; Gai, Y.A.; Wu, J. Evaluation of Angelica decursiva reference genes under various stimuli for RT-qPCR data normalization. Sci. Rep. 2021, 11, 10. [Google Scholar] [CrossRef]
  55. Ruan, W.; Lai, M. Actin, a reliable marker of internal control? Clin. Chim. Acta 2007, 385, 1–5. [Google Scholar] [CrossRef] [PubMed]
  56. Pabuayon, I.M.; Yamamoto, N.; Trinidad, J.L.; Longkumer, T.; Raorane, M.L.; Kohli, A. Reference genes for accurate gene expression analyses across different tissues, developmental stages and genotypes in rice for drought tolerance. Rice 2016, 9, 32. [Google Scholar] [CrossRef] [PubMed]
  57. Xia, S.H.; Zhang, H.; He, S.Z. Genome-Wide Identification and Expression Analysis of ACTIN Family Genes in the Sweet Potato and Its Two Diploid Relatives. Int. J. Mol. Sci. 2023, 24, 10930. [Google Scholar] [CrossRef] [PubMed]
  58. Pickart, C.M.; Eddins, M.J. Ubiquitin: Structures, functions, mechanisms. Biochim. Biophys. Acta 2004, 1695, 55–72. [Google Scholar] [CrossRef]
  59. Liu, X.; Gao, Y.; Zhao, X.; Zhang, X.; Ben, L.; Li, Z.; Dong, G.; Zhou, J.; Huang, J.; Yao, Y. Validation of Novel Reference Genes in Different Rice Plant Tissues Through Mining RNA-Seq Datasets. Plants 2023, 12, 3946. [Google Scholar] [CrossRef]
  60. Song, X.Y.; Zhang, Y.Y.; Wu, F.C.; Zhang, L. Genome-wide analysis of the maize (Zea may L.) CPP-like gene family and expression profiling under abiotic stress. Genet. Mol. Res. 2016, 15, 8023. [Google Scholar] [CrossRef]
  61. Zhao, Y.; Lang, M.; Wang, X.; Liu, G. Cloning of a DREB-like Gene RcDREB1 in Roegneria ciliars and Functional Identification of Its Protein Binding Specificity to the Dehydration-responsive Element (DRE). J. Plant Genet. Resour. 2013, 14, 857–863. [Google Scholar] [CrossRef]
  62. Inada, N. Plant actin depolymerizing factor: Actin microfilament disassembly and more. J. Plant Res. 2018, 131, 567. [Google Scholar] [CrossRef]
  63. Chanarat, S. UBL5/Hub1: An Atypical Ubiquitin-Like Protein with a Typical Role as a Stress-Responsive Regulator. Int. J. Mol. Sci. 2021, 22, 9384. [Google Scholar] [CrossRef] [PubMed]
  64. Liu, Y.; Liu, J.; Xu, L.; Lai, H.; Chen, Y.; Yang, Z.M.; Huang, B.R. Identification and Validation of Reference Genes for Seashore Paspalum Response to Abiotic Stresses. Int. J. Mol. Sci. 2017, 18, 1322. [Google Scholar] [CrossRef] [PubMed]
  65. Rasheed, R.N.; Beevy, S.S. Reliable reference gene selection for quantitative real-time PCR (qRT-PCR) in floral developmental phases of dioecious species Coccinia grandis. Gene 2024, 900, 148143. [Google Scholar] [CrossRef] [PubMed]
  66. Coito, J.L.; Rocheta, M.; Carvalho, L.; Amancio, S. Microarray-based uncovering reference genes for quantitative real time PCR in grapevine under abiotic stress. BMC Res. Notes 2012, 5, 220. [Google Scholar] [CrossRef]
  67. Cong, L.L.; Zhang, X.Q.; Yang, F.Y.; Liu, S.J.; Zhang, Y.W. Isolation of the P5CS gene from reed canary grass and its expression under salt stress. Genet. Mol. Res. 2014, 13, 9122–9133. [Google Scholar] [CrossRef]
Genes 17 00237 g001
Figure 1. Specificity of primer pairs for qRT-PCR amplification in R. ciliaris.
Figure 1. Specificity of primer pairs for qRT-PCR amplification in R. ciliaris.
Genes 17 00237 g001
Genes 17 00237 g002
Figure 2. Distribution of threshold cycle (Ct) values of candidate RGs across three sample sets. Mean Ct values with variability are shown for: (A) various tissues (leaf, root, internodes 1–3, young spike, mature spike); (B) drought stress treatments in leaves and roots; (C) all samples combined. Boxes represent the interquartile range (25th–75th percentile), with the central line indicating the median; whiskers extend from the minimum to maximum values.
Figure 2. Distribution of threshold cycle (Ct) values of candidate RGs across three sample sets. Mean Ct values with variability are shown for: (A) various tissues (leaf, root, internodes 1–3, young spike, mature spike); (B) drought stress treatments in leaves and roots; (C) all samples combined. Boxes represent the interquartile range (25th–75th percentile), with the central line indicating the median; whiskers extend from the minimum to maximum values.
Genes 17 00237 g002
Genes 17 00237 g003
Figure 3. Determination of the optimal number of RGs using GeNorm analysis. The pairwise variation (Vn/(n + 1)) was calculated to determine the minimum number of genes required for accurate normalization. The dashed line indicates the recommended cut-off value of 0.15.
Figure 3. Determination of the optimal number of RGs using GeNorm analysis. The pairwise variation (Vn/(n + 1)) was calculated to determine the minimum number of genes required for accurate normalization. The dashed line indicates the recommended cut-off value of 0.15.
Genes 17 00237 g003
Genes 17 00237 g004
Figure 4. Comprehensive ranking of the eight candidate RGs in R. ciliaris across experimental conditions (various tissues and drought stress) by RefFinder analysis. An increase in the Geomean value corresponds to a decline in gene expression stability.
Figure 4. Comprehensive ranking of the eight candidate RGs in R. ciliaris across experimental conditions (various tissues and drought stress) by RefFinder analysis. An increase in the Geomean value corresponds to a decline in gene expression stability.
Genes 17 00237 g004
Genes 17 00237 g005
Figure 5. Relative expression of RcP5CS in R. ciliaris under drought stress and across various tissues. (A) Expression profile of RcP5CS under drought stress based on RNA-seq (FPKM) data. (B–D) qRT-PCR analysis of RcP5CS expression in leaves (B) and roots (C) under drought stress, and across different tissues (D). The most stable RGs (MDH and RPL19) and the least stable genes (GAPDH and TUBA) were used as internal controls to compare the effects of RG selection on normalization. Different lowercase letters above the bars indicate significant differences among treatments or tissues at p < 0.05.
Figure 5. Relative expression of RcP5CS in R. ciliaris under drought stress and across various tissues. (A) Expression profile of RcP5CS under drought stress based on RNA-seq (FPKM) data. (B–D) qRT-PCR analysis of RcP5CS expression in leaves (B) and roots (C) under drought stress, and across different tissues (D). The most stable RGs (MDH and RPL19) and the least stable genes (GAPDH and TUBA) were used as internal controls to compare the effects of RG selection on normalization. Different lowercase letters above the bars indicate significant differences among treatments or tissues at p < 0.05.
Genes 17 00237 g005
Genes 17 00237 g006
Figure 6. Relative expression of RcPP2C in R. ciliaris under drought stress and across various tissues. (A) Expression profile of RcPP2C under drought stress based on RNA-seq (FPKM) data. (B–D) qRT-PCR analysis of RcPP2C expression in leaves (B) and roots (C) under drought stress, and across different tissues (D). The most stable RGs (MDH and RPL19) and the least stable genes (GAPDH and TUBA) were used as internal controls to compare the effects of RG selection on normalization. Different lowercase letters above the bars indicate significant differences among treatments or tissues at p < 0.05.
Figure 6. Relative expression of RcPP2C in R. ciliaris under drought stress and across various tissues. (A) Expression profile of RcPP2C under drought stress based on RNA-seq (FPKM) data. (B–D) qRT-PCR analysis of RcPP2C expression in leaves (B) and roots (C) under drought stress, and across different tissues (D). The most stable RGs (MDH and RPL19) and the least stable genes (GAPDH and TUBA) were used as internal controls to compare the effects of RG selection on normalization. Different lowercase letters above the bars indicate significant differences among treatments or tissues at p < 0.05.
Genes 17 00237 g006
Table 1. Primer sequences and amplification characteristics of candidate RGs.
Table 1. Primer sequences and amplification characteristics of candidate RGs.
Reference GenesGene NamePrimer Sequence
(from 5′ to 3′)		Amplicon
Size (bp)		E (%)R2
ADF4LActin depolymerizing factor 4 likeF: GACTTCGACTTCACCACCCC
R: AGACTGATTTCGCTGGGGTC		182108.360.9937
TUBATubulin alphaF: CCGCATCGACCACAAGTTTG
R: CATCCTCACCCTCGTCGAAC		167103.980.9970
UBL5Ubiquitin-like protein 5F: CTAGTCTCATCACACCGGCC
R: GGCAGGTCAATCACAGGAAAAG		197108.550.9949
UCE2Ubiquitin-conjugating enzymeF: CGGTCCAAGTACGAGACGAC
R: ATGGATCAGGGAGACACACG		180110.840.9908
EF1AElongation factor 1 alphaF: ACTGCCACACCTCACACATT
R: TTCTCCACGCCCTTGATGAC		244106.280.9932
MDHMalate dehydrogenaseF: TTGTTCAAGGGCTCCCGATC
R: TGTTCTGGGTGGAGACGAGA		192104.720.9914
RPL19Ribosomal protein L19F: CAGTTTGAGGCTAAGCGTGC
R: CTTTGCCTTCTTTGGTGCCG		153107.120.9908
GAPDHGlyceraldehyde-3-phophate dehydrogenaseF: GCTATCAAGGCTGCATCCGA
R: TGCTGTAACCCCACTCGTTG		184106.770.9939
Note: F and R denote the forward and reverse primers, respectively; E represents the amplification efficiency; R2 indicates the regression coefficient of the standard curve.
Table 2. Ranking of the eight candidate reference genes by expression stability across all samples, as evaluated by ΔCt, GeNorm, and NormFinder. Genes are ordered from most stable (top) to least stable (bottom) according to the M value.
Table 2. Ranking of the eight candidate reference genes by expression stability across all samples, as evaluated by ΔCt, GeNorm, and NormFinder. Genes are ordered from most stable (top) to least stable (bottom) according to the M value.
RankLeafRootInternode 1Internode 2Internode 3Young SpikeMature SpikeDrought StressTissuesAll Samples
GeneMGeneMGeneMGeneMGeneMGeneMGeneMGeneMGeneMGeneM
ΔCt analysis
1UBL5/RPL190.68RPL190.40MDH/UBL50.18RPL19/MDH0.44UBL50.14EF1A0.34RPL19/MDH0.28RPL190.98MDH1.00MDH1.06
2MDH0.73MDH0.41TUBA0.19EF1A0.48TUBA0.15UBL5/UCE20.35UCE20.29MDH0.99RPL191.05RPL191.07
3TUBA0.78UCE20.43RPL190.20ADF4L/UBL50.50MDH0.16RPL190.36EF1A0.31UCE2/
EF1A		1.12EF1A1.10EF1A1.21
4EF1A0.97TUBA0.58EF1A0.24TUBA0.59UCE2/EF1A0.17TUBA0.46UBL50.36TUBA1.40UCE21.27UCE21.30
5UCE21.71ADF4L0.65UCE20.35UCE21.39RPL190.18MDH0.69TUBA0.44UBL51.45ADF4L1.49ADF4L1.69
6ADF4L1.80EF1A0.71ADF4L0.68 ADF4L0.34ADF4L0.94ADF4L0.83ADF4L1.64TUBA1.57UBL51.80
7 UBL50.92 GAPDH1.73UBL51.70TUBA1.84
GeNorm analysis
1UBL5/RPL190.05MDH/UCE20.05TUBA/RPL190.02UBL5/ADF4L0.03MDH/EF1A0.04UBL5/RPL190.05MDH/RPL190.07MDH/RPL190.30MDH/RPL190.41MDH/RPL190.37
2MDH0.14RPL190.07UBL50.03MDH0.12UBL50.06EF1A0.14UCE20.11UCE20.55EF1A0.60EF1A0.60
3TUBA0.24ADF4L0.28MDH0.04EF1A0.18RPL190.10UCE20.15EF1A0.13EF1A0.66UCE20.80UCE20.79
4EF1A0.39TUBA0.38EF1A0.08RPL190.22TUBA0.11TUBA0.21UBL50.17TUBA0.78ADF4L1.03ADF4L1.08
5UCE20.75EF1A0.45UCE20.14TUBA0.31UCE20.12MDH0.33TUBA0.23UBL50.95TUBA1.15UBL51.26
6ADF4L1.05UBL50.59ADF4L0.29UCE20.62ADF4L0.19ADF4L0.50ADF4L0.40ADF4L1.16UBL51.31TUBA1.42
7 GAPDH1.31
NormFinder analysis
1UBL5/RPL190.02MDH/UCE20.026MDH/UBL50.01RPL190.10UBL50.03UBL5/RPL190.03MDH0.03MDH0.26MDH0.07RPL190.08
2MDH0.08RPL190.03TUBA0.03TUBA0.14TUBA0.04EF1A/UCE20.05RPL190.04RPL190.28RPL190.44MDH0.18
3TUBA0.13TUBA0.45EF1A0.06MDH0.16UCE20.08TUBA0.33UCE20.06UCE20.59EF1A0.53EF1A0.61
4EF1A0.58ADF4L0.48RPL190.07EF1A0.23MDH0.11MDH0.63EF1A0.18EF1A0.66UCE20.85UCE20.76
5UCE21.65EF1A0.65UCE20.32UBL50.41EF1A0.13ADF4L0.92UBL50.19TUBA1.14ADF4L1.22ADF4L1.38
6ADF4L1.75UBL50.88ADF4L0.68ADF4L0.42RPL190.14 TUBA0.39UBL51.17TUBA1.33UBL51.60
7 UCE21.38ADF4L0.38 ADF4L0.82ADF4L1.38UBL51.53TUBA1.65
8 GAPDH1.50
Note: The M values for GAPDH are only presented for the drought stress dataset, as its expression was undetectable in tissues other than leaves and roots.
Table 3. Stability evaluation of the eight candidate reference genes by BestKeeper analysis. Genes are ordered from most stable (top) to least stable (bottom) based on the combined standard deviation (SD) and coefficient of variation (CV). Lower values for both parameters indicate greater expression stability.
Table 3. Stability evaluation of the eight candidate reference genes by BestKeeper analysis. Genes are ordered from most stable (top) to least stable (bottom) based on the combined standard deviation (SD) and coefficient of variation (CV). Lower values for both parameters indicate greater expression stability.
RankLeafRootInternode 1Internode 2Internode 3Young SpikeMature Spike
GeneSDCVGeneSDCVGeneSDCVGeneSDCVGeneSDCVGeneSDCVGeneSDCV
1TUBA0.261.17UCE20.140.74UCE20.080.38UBL50.030.15UBL50.150.63UBL50.100.46ADF4L0.090.42
2EF1A0.381.88MDH0.180.94RPL190.100.46ADF4L0.040.19MDH0.160.82RPL190.110.56UBL50.431.94
3RPL190.381.84RPL190.221.15TUBA0.110.65MDH0.090.46EF1A0.191.05UCE20.231.14UCE20.452.13
4UBL50.381.90UBL50.401.92UBL50.130.60EF1A0.191.07TUBA0.191.06EF1A0.231.36RPL190.462.29
5MDH0.412.07TUBA0.482.53MDH0.130.65RPL190.221.08ADF4L0.241.08TUBA0.382.16MDH0.522.66
6ADF4L0.893.58ADF4L0.502.08EF1A0.221.15TUBA0.422.34RPL190.251.24MDH0.452.37EF1A0.583.13
7UCE2 *1.396.48EF1A0.532.99.ADF4L0.632.70UCE2 *1.185.56UCE20.251.18ADF4L0.472.22TUBA0.683.49
RankDrought stressTissuesAll samples
GeneSDCVGeneSDCVGeneSDCV
1UBL50.562.74MDH0.492.54MDH0.623.13
2MDH0.603.01RPL190.663.29RPL190.733.63
3UCE20.743.68UCE20.733.56UCE20.743.65
4RPL190.753.70UBL50.904.13UBL50.924.38
5EF1A0.995.18EF1A0.914.95EF1A *1.035.50
6ADF4L *1.185.20ADF4L *1.034.48ADF4L *1.114.88
7TUBA *1.426.81TUBA *1.236.52TUBA *1.728.63
8GAPDH *1.498.29
Note: * indicates excluded gene (SD > 1).
Table 4. Identification of optimal single and paired reference genes for R. ciliaris across various tissues and under drought stress conditions in leaves and roots.
Table 4. Identification of optimal single and paired reference genes for R. ciliaris across various tissues and under drought stress conditions in leaves and roots.
Experimental ConditionsSingle Most Stable Reference GenesOptimal Combination Reference Genes
Different tissuesMDHMDH + RPL19
Drought stressMDHMDH + RPL19
All samplesMDHMDH + RPL19
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Luo, Q.; Liu, Y.; Wang, Y.; Zhang, G.; Liu, J.; Li, H.; Liang, Z.; Liu, Y.; Bai, L.; Liu, S. Transcriptome-Based Selection and Validation of Reference Genes for Gene Expression Analysis in Roegneria ciliaris ‘Liao Sheng’ Across Various Tissues and Under Drought Stress. Genes 2026, 17, 237. https://doi.org/10.3390/genes17020237

AMA Style

Luo Q, Liu Y, Wang Y, Zhang G, Liu J, Li H, Liang Z, Liu Y, Bai L, Liu S. Transcriptome-Based Selection and Validation of Reference Genes for Gene Expression Analysis in Roegneria ciliaris ‘Liao Sheng’ Across Various Tissues and Under Drought Stress. Genes. 2026; 17(2):237. https://doi.org/10.3390/genes17020237

Chicago/Turabian Style

Luo, Qianyun, Yue Liu, Yifan Wang, Guanghao Zhang, Jiafen Liu, Hongxin Li, Zhen Liang, Ying Liu, Long Bai, and Sijia Liu. 2026. "Transcriptome-Based Selection and Validation of Reference Genes for Gene Expression Analysis in Roegneria ciliaris ‘Liao Sheng’ Across Various Tissues and Under Drought Stress" Genes 17, no. 2: 237. https://doi.org/10.3390/genes17020237

APA Style

Luo, Q., Liu, Y., Wang, Y., Zhang, G., Liu, J., Li, H., Liang, Z., Liu, Y., Bai, L., & Liu, S. (2026). Transcriptome-Based Selection and Validation of Reference Genes for Gene Expression Analysis in Roegneria ciliaris ‘Liao Sheng’ Across Various Tissues and Under Drought Stress. Genes, 17(2), 237. https://doi.org/10.3390/genes17020237

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop