Introduction

The genus Ribes L. is the sole genus of Grossulariaceae or Ribesiaceae1,2and in some studies, this genus has been classified in Saxifragaceae3,4. Ribes species are evergreen or semi-evergreen shrubs, mainly distributed in mild-climate regions, such as North America, Eurasia, and the Andes Mountains5. China is one of the distribution centers of Ribes, with about 59 species and 30 varieties, mainly distributed in the southwest, northwest, and northeast. It not only has rich species diversity, but also includes the main stages of phylogenetic groups6,7.

Ribes plants have important economic value due to their rich secondary compounds. The fruits contain various abundant vitamins, sugars, and organic acids, and have important edible value8. In addition, the roots and seeds of certain species can be used for medicine due to their rich contents of flavonoids, proanthocyanidins, phenolglycosides and other substances9,10. Additionally, some species have beautiful leaf shapes, tree shapes, and vibrant colors, and can be used as ornamental and landscaping trees11.

Ribes odoratum H.L.Wendl. is native to North America, distributes along rivers in mountains12. It was introduced and planted in Liaoning, Heilongjiang and other places in China. In 1998, R. odoratum was introduced to Qinghai province. After nearly 20 years of domestication and cultivation, it has been found that R. odoratum has strong adaptability to the arid climate of the Qinghai Plateau. Therefore, the study of the adaptation mechanism to drought stress conditions and further exploration of genes that are related to its drought stress adaptation are key steps to subsequent breeding.

For the research about gene function, reference gene selection is a key factor because the stability of a reference gene could influence the accuracy of results. In general, most of the research uses housekeeping genes as the reference, such as Actin isoform B(ACTB), 18 S rRNA, GAPDH, EF1 and some others13,14. In the previous researches about Ribes plants, ACTIN was used as a reference gene15,16,17.

However, in some reports, housekeeping genes acting as reference genes could be affected by many factors, such as plant development stages, stress conditions, tissue types and experimental factors, and be differently expressed in different conditions18,19. For example, in a study conducted in 2017, the expression stability of potential housekeeping genes at four different developmental stages (immature, green-ripe, breaker, and red-ripe) of tomatoes was verified. The results showed that the cycle threshold (Ct value) of different genes varied at each developmental stage20. In this study, the cells of Hypericum perforatum were treated with silver nanoparticles and gold nanoparticles, and the expression stability of multiple housekeeping genes under different treatment conditions and time points was analyzed. It was found that there were differences in the expression stability of these genes under different nanoparticle types and treatment durations21. Studies on the stability of reference genes in Hypericum perforatum under cold and heat stresses have shown that their expressions also vary under different stress treatments22. Randomly adopted reference genes could affect the result accuracy, lead to deviation, and finally draw a wrong conclusion23. Therefore, stably expressed reference genes suitable for experimental requirements are a prerequisite for accurate results and conclusions.

To date, there have been no reports on specific reference genes suitable for Ribes odoratum, especially under drought stress conditions. Building on recent research advances, this study screened reference genes through transcriptome data analysis and evaluated their stability under drought stress, aiming to identify the optimal reference genes for gene expression normalization in Ribes odoratum under drought stress. This will provide standardized tools for subsequent gene expression analyses related to drought resistance.

Materials and methods

Stress treatment and collection

The plant material used in this study is Ribes odoratum H.L.Wendl. It was collected on 19 May 2022 in Qinghai, China, and was formally identified by Dr. Zhang Defang (co-author). The voucher specimen of this plant, identified in 2016, has been deposited in the herbarium of the Northwest Institute of Plateau Biology, Chinese Academy of Sciences, with the specimen deposit number HNWP00016174. All of the plants were treated at Forestry Experimental Base of Qinghai University in Qinghai Province in July 2023. A total of 18 samples were treated, six group were set up (CK, T1, T2, T3, T4 and T5) with 5 kg soil in each pot of every plant. In the first round of experiments, all of treatment sample were watered 500 ml of water simultaneously watering on the 5 treatment groups and CK. From then on, keep the soil surface moist throughout the entire experimental process for the CK, watering 500 ml every three days for T1, 500 ml every six days for T2, 500 ml every nine days for T3, 500 ml every twelve days for T4 and 500 ml every fifteen days for T5 intervals respectively for three times. Leaf samples were collected on the day after the 3rd treatment for each group, and phenotypic observations were recorded. All samples were immediately frozen in liquid nitrogen for RNA extraction. Phenotypic manifestations under drought stress are shown in Fig. 1.

Fig. 1
figure 1

The phenotype of Ribes odoratum under drought stress.

Full size image

Transcriptome sequencing, date analyzing and Candidate reference gene screening

Total RNA extraction, assessment and verification were carried out using TaKaRa MiniBEST Plant RNA Extraction Kit (TaKaRa, Dalian, China), Nano 6000 Assay Kit of the Bioanalyzer 2100 system (Agilent Technologies, CA, USA) and Nanodrop One (Thermo Fisher Scientific, Inc., USA) respectively.

The construction and assessment of Sequencing libraries, the processing of data and the analysis of gene expression levels were adapted from previous research24.

Candidate reference gene selecting and primer design

Candidate genes were selected under different stresses according to the FPKM values and based on the CV and SD values of the FPKM, with the thresholds set at SD < 3.6 and CV < 0.55. The primers of candidate genes were designed using primer 3 plus website(https://www.primer3plus.com/index.html). Two primer pairs were designed for each gene and used for subsequent qPCR amplification.

The cDNA synthesis, primer selection and specificity validation based on qPCR

cDNA was synthesized using the FastKing gDNA Dispelling RT SuperMix (TIANGEN, China) with 500 ng of RNA template per reaction, following the manufacturer’s recommended protocol and thermocycling conditions. All primers were validated by qPCR, and experimental conditions were adapted from previous studies24.

The primers that could amplify single, specific, clear and expected size bands were selected for subsequent Realtime-PCR. The reaction mixture contained 10 µL of TB Green® Premix Ex TaqTM (Tli RNaseH Plus), 0.6 µL of each forward and reverse primer, and 9 µL of nuclease-free water, bringing the total volume to 20 µL. Each sample was prepared in triplicate technical replicates (n = 3) to assess intra-assay precision.

Amplification was standardized in a CFX Connect Real-Time System (BIO-RAD) with 95 ℃ for 30 s, 95 ℃ for 5 s, and 60 ℃ for 30 s by 50 cycles. The specificity of all amplification products was verified by melt curve analysis, with a single peak indicating primer specificity. A standard curve was generated using 10-fold serially diluted pooled cDNA, and the amplification correlation coefficient (R²) and efficiency (E) of each primer were determined based on this curve. The amplification efficiency (E) was calculated using the following formula: E = [10(−1/slope) -1]×100%25.

Data analysis

The stability of data analysis for all candidate reference genes was assessed using geNorm, NormFinder, BestKeeper, and Delta Ct algorithms in the RefFinder platform accessed on 25 August 2024.

The RefFinder platform automatically calculates the geometric mean of the weights of each gene’s CT values across the four algorithms. Subsequently, all candidate reference genes were ranked based on the geometric mean of each gene and generated a comprehensive ranking. The gene with the lowest geometric mean was considered a stable reference gene, while the gene with a highest geometric mean is considered a less stable gene26.

The Venn diagrams were drawn on the website (https://jvenn.toulouse.inrae.fr/app/exampl e.html).

Result

Sequencing result

A total of 115.9G raw data were obtained from 18 samples, and each sample data was 6.44G on average. The T4_1 sample generated the least amount of data (Table S1). The average Q20 value reached 97.31%, and the average Q30 value reached 92.88%. All raw data can be found in the National Center for Biotechnology Information (NCBI) SRA database (Accession number: PRJNA1171611).

Expression profiles of candidate genes in transcriptome sequencing

A total of 35 candidate reference genes were obtained from transcriptome sequencing according to the SD and CV values. As Table 1 shows, the average FPKM value ranged from 2.274 to 77.098, and the SD value, which ranged from 0.104 to 3.550, increased with the average FPKM value, while the CV values, which ranged from 0.026 to 0.054, not showed same trends.

Table 1 Expression profiles of candidate reference genes in transcriptome sequencing.
Full size table

Specificity and efficiency of candidate genes

Seventeen genes were screened by 1.5% agarose gel electrophoresis, which could produce clear, specific bands of expected size without primer dimers. The correlation coefficients of these primers ranged from 0.974 to 0.999, and the amplification efficiency ranged from 95.3 to 109.9%, making them suitable for subsequent qPCR experiments (Table 2). The qPCR results showed that the melting curves and amplification curves of the 17 candidate genes are presented in Figs. 2 and 3, respectively, indicating that all primers have good amplification specificity. Except for the genes Cluster-37.22273 and Cluster-37.22897, the melting curves of the other genes all showed a single main peak. The abnormality of these two genes may be due to operational issues (Table 3).

Table 2 The information of candidate internal reference genes for qPCR.
Full size table
Fig. 2
figure 2

The amplification curve of all candidate reference genes in qPCR.

Full size image
Fig. 3
figure 3

The melting curve of all candidate reference genes in qPCR.

Full size image

Stability evaluation

In the GeNorm analysis, the average expression stability (M value) of the 17 candidate reference genes ranged from Cluster-37.23244 (0.389) to Cluster-37.32890 (3.599). The top five most stable genes were Cluster-37.23244(0.389), Cluster-37.24426 (0.389), Cluster-37.24001 (0.515), Cluster-37.22646 (0.596), and Cluster-37.22228 (0.706), among which Cluster-37.22228 was also ranked within the top five in the other three methods(Table 3).

In the Normfinder analysis, stability values ranged from Cluster-37.22228 (0.277) to Cluster-37.32890 (11.740). The top five most stable genes were Cluster-37.22228 (0.277), Cluster-37.22897 (0.407), Cluster-37.22646 (0.734), Cluster-37.22273 (0.853), and Cluster-37.24280 (1.051). Conversely, Cluster-37.23838 (4.218), Cluster-37.24007 (4.369), and Cluster-37.32890 (11.740) exhibited the poorest stability, with all evaluation values exceeding 4 (Table 3).

Unlike the first two algorithms, the BestKeeper analysis assessed gene stability by the standard deviation (SD) of Ct values. Cluster-37.22897 (0.53), Cluster-37.22066 (0.64), Cluster-37.22273 (0.85), Cluster-37.22228 (0.91), and Cluster-37.24280 (1.22) ranked among the top five in this analysis, with the lowest SD values indicating the highest stability. Cluster-37.32890 ranked last with the highest SD value (8.01), reflecting the poorest stability. In the Delta CT analysis, stability was evaluated based on the SD of candidate genes’ Delta Ct values. The top five stable genes were Cluster-37.22228 (2.38), Cluster-37.24280 (2.57), Cluster-37.22066 (2.99), Cluster-37.22897 (2.51), and Cluster-37.24426 (2.53). By contrast, Cluster-37.23838 and Cluster-37.32890 had the largest SD values (both > 4), indicating the poorest stability assessments (Table 3).

Based on the comprehensive ranking, the top five genes in order were Cluster-37.22228, Cluster-37.22646, Cluster-37.22897, Cluster-37.23244, and Cluster-37.24426 (Table 3).

Table 3 The expression stability of candidate reference genes from three software analyses.
Full size table

As shown in Fig. 4, a Venn diagram was constructed based on the top 10 genes from each of the four methods. The results showed that six genes—Cluster-37.22228, Cluster-37.22646, Cluster-37.22897, Cluster-37.24001, Cluster-37.22273, and Cluster-37.24280—were jointly identified as stable candidate reference genes by all four methods and ranked according to their stability. Cluster-37.24426 and Cluster-37.23244 were the two genes ranked within the top 10 exclusively in the geNorm algorithm. Cluster-37.25912 was a candidate gene shared by the NormFinder and BestKeeper algorithms. Both Cluster-37.26541 and Cluster-37.24801 ranked among the top 10 genes in both geNorm and Delta Ct analyses.

Fig. 4
figure 4

The Veen diagram of candidate genes.

Full size image

Discussion

With the deteriorating natural environment, stress-tolerant plants have become the focus of agricultural researchers. In particular, the research on genes related to drought resistance, salt-alkali resistance, etc. is of utmost importance for stress-tolerant plants. The development and evaluation of the most stable reference genes form the basis for uncovering key stress-tolerance genes27,28. Therefore, we conducted the development and testing of reference genes for R. odoratum under different drought stress conditions. The results are as follows:

Cluster-37.22228 was annotated as the homeobox-DDT domain-containing protein RLT2, which has a homeobox domain. It plays a significance role in regulating gene expression patterns under multiple non-biological stress factors29. For instance, in the research on the response variations of mustard seedlings to heat stress, it was discovered that the homeobox-DDT domain-containing protein RLT2 participates in the gene expression of the heat stress response, which helps plants adapt to high-temperature environments30. RLT2, along with RLT1/2, is components of the ISWI chromatin remodeling complex (CRAF type) in Arabidopsis, and they maintain the structural stability of the complex, influence floral morphological development31,32regulate flowering time33and is related to chromatin remodeling complexes34.

Therefore, we speculate that Cluster-37.22228, which is stably expressed at different drought stress levels, may be related to drought tolerance and adaptation.

Cluster-37.22646 was annotated as homogentisate solanesyltransferase (HST), which is a key enzyme in the plastoquinone (PQ) biosynthesis pathway and affects PQ synthesis. The catalytic function of HST in the PQ biosynthesis pathway is crucial for maintaining normal photosynthesis and gene expression regulation in plants, and it is closely related with the environmental adaptability in Arabidopsis thaliana35. It is also related to plant flowering time36,37. For example, in Arabidopsis thaliana, the disruption of HST can lead to an albino and dwarf phenotype, as well as defects in roots, trichomes, and stomata38. Its stable expression in this research indicates HST’s role in maintain normal physiological activities in response to drought stress of Ribes.

According to the annotation, Cluster-37.22897 is a GATA transcription factor. In plants, GATA transcription factors can be divided into four categories39specifically Class A, Class B, Class C, and Class D, each presenting distinctive characteristics in both structure and function. For example, Class A plays a role in photomorphogenesis and relates to drought stress response in tomatoes40,41. Class B, especially the AtGNC and AtGNL, is involved in the light-dependent chlorophyll synthesis process42. Based on these functions in other plants, we believe the Cluster-37.22897 is related to drought stress or regulate some key gene expression.

Cluster-37.24001 was annotated as E3 ubiquitin-protein ligase ARI5. The E3 ubiquitin ligase TaSDIR1-4 A in wheat is associated with the drought resistance in wheat43. The E3 ubiquitin ligase PUB44 can activate cellular adaptive physiological activities in response to osmotic stress44. Although there have been no reports on E3 ubiquitin-protein ligase ARI5 in R. odoratum, according to previous research, we suspect Cluster-37.24001 is related to the adaptation to drought stress in R. odoratum.

Cluster-37.22273 was annotated as ubiquitin carboxyl-terminal hydrolase 3 (UCH3). UCH3 influences multiple aspects of plant growth and development. In Arabidopsis thaliana, it plays a vital role in regulating the circadian clock of Arabidopsis under high-temperature conditions and also affects the flowering time of plants45. However, its function in R. odoratum remains to be explored.

Cluster-37.24280 was annotated as U-box domain-containing protein 62, which belongs to the U-box gene family. Previous studies have shown that the U-box gene family plays significant roles in protein modification46,47plant growth and development, as well as in response to environmental stresses48. Therefore, we think its function in the R. odoratum may be related to drought stress adaptation.

Cluster-37.23244 was annotated as E3 ubiquitin-protein ligase UPL3. UPL3, as a specific type of E3 ubiquitin-protein ligase, plays a crucial role in plant growth and development. It is particularly involved in the regulation of gene expression related to seed size, lipid content during seed maturation, and crop yield49. Additionally, UPL3 also plays a key role in promoting the processing of immune-related proteasomes50.

Cluster-37.24426 was annotated as NF-X1 type zinc finger protein NFXL1. The NFXL1 protein has a positive impact on adaptation to salt stress and drought stress51,52.

In this study, we selected 17 candidate reference genes for qPCR analysis of seedlings in Ribes odoratum. under different drought stress. Cluster-37.22228, Cluster-37.22646 and Cluster-37.22897 were the top three stable reference genes, and the best reference genes that could be used as internal controls in the Ribes odoratum was Cluster-37.22228 and Cluster-37.22646. In conclusion, this study provides an important basis for quantifying reference gene expression levels in Ribes odoratum.