Introduction

Potato blackleg disease caused by Pectobacterium atrosepticum is a severe bacterial disease that can affect potatoes throughout the entire cultivation of potato1. This bacterium primarily infects plant stems and tubers of the potato. In the early stages of infection, typical symptoms include stunted seeding growth, shortened internodes, and wrinkling and curling of leaves, with characteristic blackening of the tissue at the base of the stem2. As the disease progresses, the vascular tissues become obstructed, leading to systemic wilting and ultimately resulting in plant death. During storage, the disease also causes extensive rot of tubers, severely impacting the yield and quality of potatoes3,4. Therefore, identifying and utilizing resistance genes to P. atrosepticum in potatoes may hold significant implications for potato production.

The analysis of disease resistance-related gene expression patterns in plant-pathogen interactions is highly dependent on molecular techniques such as real-time fluorescence quantitative PCR (RT-qPCR)5,6. RT-qPCR is recognized for its efficiency, accuracy, cost-effectiveness, and user-friendliness, making it a widely utilized method for analyzing resistance gene expression. It can effectively detect RNA in a wide concentration range, encompassing both high and low-abundance transcripts. In addition, RT-qPCR can be applied to types of samples, including tissues, cells, and body fluids. However, it is important to note that RT-qPCR results effectively maintain a constant expression level across treatments and replicates7,8. Therefore, the selection of stably expressed reference genes is critically important in this context. These genes serve as a stable molecular reference standard, enabling the precise validation of the expression levels of key genes following pathogen infection, which is essential for assessing their potential impact on disease resistance9.

Reference genes are a category of genes found in all cells, playing a vital role in maintaining normal cellular structure and function. These genes encode proteins essential for cell survival and participate in fundamental metabolic processes crucial for cell viability. They usually show consistent and stable expression levels in different cell types and tissues, render them valuable as reference genes in gene expression studies10,11. Commonly utilized internal reference genes in research include GAPDH, which is a key enzyme in the glycolytic pathway responsible for catalyzing the production of glycerol-3-phosphate from phosphatidyl glycosides; Actin, a crucial protein that forming the cytoskeleton, providing structural support and facilitates cellular movement and material transport; 18S rRNA, a component of the small subunit ribosomal RNA involved in reading messenger RNA and protein synthesis; 28S rRNA, part of the large subunit ribosomal RNA responsible for peptide bond formation; Ubiquitin, which is involved in protein regulation and degradation; Tubulin, a key protein in the cytoskeleton and cell division processes12,13,14. In theory, reference genes are expected to be conserved and stable. However, numerous studies have indicated that the nature and condition of the samples can influence the expression of reference genes. For example, in response to cold stress in potatoes, the genes EF1α and L2 demonstrated good stability, while Actin and Tubulin exhibited lower stability15. Similarly, under drought stress, sec3, and L8 were found to be more stable, whereas APRT and Tubulin showed less stability16. Furthermore, there are significant differences in the stability of amplification products of the same reference gene between different species. For instance, in Nicotiana benthamiana, EF1-α1 and EF1-α2 displayed the lowest stability after salt stress, whereas 18rRNA and GAPDH showed the highest stability17. Furthermore, notable differences can be observed between different primer amplification products of the same gene. For example, in salt-stressed pear trees, the stability of TUB-A, TUB-B, EF1α-2A, and EF1α-2B varies, with EF1α-2A showing significantly higher stability than EF1α-2B and TUB-B exhibiting greater stability than TUB-A18.

This study aims to systematically assess the expression stability of seven candidate reference genes of potato infected by P. atrosepticum using RT-qPCR, and select the most suitable reference genes with geNorm, NormFinder, BestKeeper, ΔCt, and RefFinder algorithms methods, which provide a standardized reference for the subsequent studies on screening of disease resistance genes and exploration of disease resistance mechanisms in potato—P. atrosepticum interactions.

Materials and methods

Materials

Cultivation of test strains and sampling

The experimental bacterial strain P. atrosepticum and two potato cultivars—the disease-resistant genotype Qingshu No. 9 and the susceptible genotype Leshu No. 1 were provided by the Laboratory of Integrated Pest Management in Agriculture, Qinghai Province, China (Table 1), the disease resistance of the two potato varieties was identified based on our research group’s previous comprehensive evaluation and classification of potato cultivars for resistance to P. atrosepticum.

Table 1 Sources and characteristics of different potato varieties.
Full size table

The pathogenic P. atrosepticum was revived from low-temperature storage and cultured on an LB medium for 48 h. Single bacterial colonies were selected and inoculated into 100 mL of sterilized LB medium. The cultures were incubated at 28 °C with shaking at 180 rpm for 24 h to reach a bacterial suspension concentration of 106 cfu/mL. Different varieties of potato tubers were planted in 150 mm × 12.5 mm pots in a greenhouse. After the potato plants had grown to a height of approximately 10–15 cm, 10 μL of the 106 cfu/mL bacterial suspension was injected into the above-ground stem base of the potato plants in the treatment group, while the control group was inoculated with an equal volume of sterile water. Each treatment was replicated in triplicate, Samples were taken 1, 3, and 5 dai after inoculation (dai). The tissues at the junction of disease and health areas at the inoculation site were frozen with liquid nitrogen and stored at − 80 °C for RNA extraction. Some of the samples were used for transcriptome sequencing, and some were utilized for this experiment.

Experimental methods

RNA extraction and qPCR reagents

Total RNA was extracted from all biological replicates of two potato cultivars at different time points 1, 3, and 5 days after inoculation (dai), following independent processing after grinding in liquid nitrogen. For each sample, 0.1 g of tissue was weighed, and total RNA was isolated according to the manufacturer’s provided with the RNA extraction kit (Tiangen Biotech Co., Ltd., Polysaccharide Polyphenol Plant Total RNA Extraction Kit, Cat. No. DP441). The concentration and purity of the extracted RNA were determined using a micro nucleic acid analyzer. Subsequently, the RNA was diluted to a concentration of 100 ng/μL, and 1 μL of total RNA from each sample was used as a template for cDNA synthesis using the FastKing cDNA Synthesis Kit (Tiangen Biotech Co., Ltd., Cat. No. KR118). Both RNA extraction and cDNA synthesis were performed independently for each biological replicate. Finally, RT-qPCR analysis was conducted using the FastReal Fluorescent Quantitative PCR Premix SYBR Green Kit (Tiangen Biotech Co., Ltd., Cat. No. FP217-01), with each RT-qPCR reaction carried out in triplicate to ensure technical reproducibility.

Design of candidate reference gene primers

Candidate reference genes for potatoes were searched through the National Genomic Data Center website (https://ngdc.cncb.ac.cn/icg/species/accession/ICG00103) and relevant literature. Combined with transcriptomic data from two potato varieties infected with the pathogen P. atrosepticum, seven genes were ultimately selected as reference genes: EF1α, Tubulin, eIF5A3, Actin, CYP3, GAPDH, and Ubiquitin. Using the gene sequences obtained from transcriptome sequencing, specific primers for fluorescent quantification were designed with Primer Premier 5 software and synthesized by Shanghai Sangon Biotech Co., Ltd. (Table 2), the primers and corresponding gene sequences of the seven candidate internal reference genes have been stored in the NCBI GeneBank database with entry numbers PP741381-PP741387. The list of seven candidate reference genes and their corresponding FPKM expression values derived from RNA-seq date from P. atrosepticum inoculated potato experiment is described in supplementary Table S1.

Table 2 Primer sequences used to amplify candidate reference genes.
Full size table

Construction of qRT-PCR standard curve and amplification efficiency analysis

The reaction volume for qRT-PCR was 25 μL, with components including 12.5 μL 2 × FastReal qPCR PreMix (SYBR Green), 0.75 μL each of forward and reverse primers, 1 μL 50 × ROX Reference Dye, 1 μL cDNA template, and 7 μL RNase-Free ddH2O. The amplification program included of an initial denaturation step at 95 °C for 2 min, followed by 40 cycles of denaturation at 95 °C for 5 s and annealing/extension at 60 °C for 15 s.

To ensure that the amplification efficiencies of all reference genes within the same reaction system, cDNA from samples of two varieties treated at different time points was mixed in equal amounts. The pooled cDNA was underwent a fivefold dilution series (5–1, 5–2, 5–3, 5–4, and 5−5 fold dilutions of the original cDNAs were used) for qRT-PCR amplification. Standard curves were generated for the seven reference genes. The amplification efficiency (E) of the primers was determined by calculating the slope of the standard curve and applying the equation E = (10–1/slope − 1) × 100%. Primers with an amplification efficiency between 95 and 110% were considered acceptable.

Stability assessment of reference gene amplification products

The CT values, which reflect reflecting the expression abundance of reference genes across samples were processed using Microsoft Excel 2019. According to the software instructions, geNorm, NormFinder, BestKeeper, ∆Ct, and Refinder software were used to analyze the expression stability of reference genes under pathogen stress. Based on the results obtained from the software, the most stable reference genes were selected.

Results

Assessment of total RNA quality

Total RNA was extracted from the samples of two potato varieties and analyzed by 2% agarose gel electrophoresis. The results of the agarose gel electrophoresis analysis (Supplementary Fig. S1a and S1b) showed that the 28S and 18S rRNA bands in both samples were identifiable and well-defined, indicating that no significant RNA degradation had occurred. The nucleic acid purity ratios for all samples were provided in Supplementary Table S2.

Candidate reference gene primer specificity and amplification efficiency

Candidate reference gene primer sequences, ranging in size from 104 to 207 bp, were utilized in conventional PCR amplifications on cDNA templates from all samples encompassing two potato varieties. The RT-qPCR amplifications yielded single bands (Supplementary Fig. S2) of expected product lengths. Subsequent RT-qPCR analysis revealed that seven primer pairs displayed melting curves with a single peak signal, indicating the absence of nonspecific amplification (Fig. 1). Amplification efficiency was calculated from the standard curves of the amplified fragments, with all seven primers demonstrating efficiencies exceeding 100% and a coefficient of determination (R2) exceeding 0.990 (Supplementary Table S3). These findings suggest that the designed primers exhibit good specificity and are suitable for further experiments.

Fig. 1
Fig. 1The alternative text for this image may have been generated using AI.
Full size image

Melt curves of the seven candidate reference gene primers: (a) Actin, (b) CYP3, (c) EF1α, (d) GAPDH, (e) eIF5A3, (f) Tubulin and (g) Ubiquitin.

Candidate reference gene Ct value expression

The Ct value is a crucial parameter in real-time quantitative fluorescence reactions for measuring target DNA molecules. Amplification results from RT-qPCR indicate that the expression levels of EF1α, eIF5A3, Tubulin, Ubiquitin, GAPDH, Actin, and CYP3 in samples from two distinct potato cultivars ranged from 16.8–18.20, 19.58–22.41, 18.25–20.54, 23.39–27.55, 18.85–21.63, 20.16–22.90, and 21.81–24.94, respectively. According to the Ct values (Fig. 2), EF1α amplification products exhibited the most stable Ct values with the lowest variability, a characteristic typically associated with constitutively high expression levels. Conversely Ubiquitin amplification products displayed the highest variability in Ct values, a pattern typically observed for genes with relatively low expression levels.

Fig. 2
Fig. 2The alternative text for this image may have been generated using AI.
Full size image

Distribution of Ct for a potato- P. atrosepticum interaction experiment values of seven candidate internal reference genes.

Expression stability of candidate internal reference genes

GeNorm analysis

GeNorm software is a widely used algorithm that assesses gene stability by calculating variability in gene expression levels across samples. This program calculates a stability value (M value) for each candidate reference gene, where lower M values correspond to higher expression consistency across diverse biological conditions and treatment regimens19. Figure 3 presents the expression stability M values of candidate reference genes based on GeNorm analysis. The results show that the M values of all candidate genes are below the recommended threshold of 1.5, indicating stable expression of all seven reference genes. GAPDH, Actin, and EF1α are the most stable reference genes in potato tissues at 1, 3, and 5 days after inoculation (dai) with P. atrosepticum, respectively. Moreover, the analysis of expression stability in all samples showed that eIF5A3 and Actin were the most stable reference genes, while Ubiquitin was the least stable.

Fig. 3
Fig. 3The alternative text for this image may have been generated using AI.
Full size image

Analysis of the expression stability of candidate reference genes based on GeNorm.

To achieve more reliable evaluation results, the optimal number of reference genes was determined through pairwise variation analysis of normalization factors (Vn/Vn+1). When Vn/Vn+1 is less than 0.15, it indicates that reference genes are sufficient for stable normalization. Figure 4 shows the analysis of the optimal number of reference genes under different treatments based on GeNorm. Following infection with P. atrosepticum, the V2/3 values at 1, 3, 5 dai, and across all samples are 0.0748, 0.0732, 0.09, and 0.121, respectively. These results demonstrate that the incorporation of two reference genes provides adequate normalization stability for reliable RT-qPCR.

Fig. 4
Fig. 4The alternative text for this image may have been generated using AI.
Full size image

Pairwise variation analysis of candidate reference genes: Vn/n + 1 < 0.15 indicates that the number of suitable genes is n, while Vn/n + 1 > 0.15 indicates that the number of suitable genes is n + 1.

BestKeeper analysis

BestKeeper assesses reference gene stability by calculating the standard deviation (SD) and coefficient of variation (CV) of gene expression levels. Lower SD values indicate higher stability, whereas genes with an SD exceeding a predefined threshold are considered unstable20. According to the BestKeeper analysis results (Table 3), EF1α shows the highest stability, followed by Tubulin and eIF5A3 with good stability, while Ubiquitin and GAPDH exhibit poor stability. EF1α has the smallest standard deviation and coefficient of variation, indicating highly stable expression levels across all experimental conditions. These findings suggest that EF1α is the most reliable reference gene for normalization in this study.

Table 3 The rank of reference genes for normalization calculated by the BestKeeper program.
Full size table

NormFinder analysis

The stability of candidate reference genes was evaluated using NormFinder, which ranks genes based on their stability value (SV), where a lower SV indicates higher gene stability21,22. As shown in Table 4, there was significant variability in the stability of the seven candidate reference genes. EF1α and Tubulin were the most stable genes in 1 dai samples, while CYP3 and UB were the least stable. CYP3 and Actin were the most stable at 3 dai, with Ubiquitin and eIF5A3 were the least stable. At 5 dai, EF1α and eIF5A3 were the most stable, whereas GAPDH and Tubulin were the least stable. Across all samples, EF1α and Tubulin were the most stable genes, with GAPDH and Ubiquitin were the least stable. Furthermore, discrepancies were observed between the stability rankings generated by NormFinder and those derived from GeNorm analysis. For instance, while eIF5A3 and Actin were identified as the most stable reference genes in all samples by GeNorm analysis, in NormFinder, they are ranked third and fourth, respectively.

Table 4 Shows that the Expression stability of candidate reference genes was analyzed by NormFinder, ∆Ct, and RefFinder software.
Full size table

∆Ct analysis

ΔCt analysis ranks candidate reference genes by average standard deviation, with lower average standard deviations indicating greater stability of the reference genes. Stability analysis of seven candidate reference genes reveals that EF1α was the most stable gene at 1 dai, with Ubiquitin being the least stable; eIF5A3 was the most stable at 3 dai, with Actin being the least stable; and Actin was the most stable at 5 dai, with Tubulin being the least stable. An analysis of all samples shows that EF1α was the most stable gene overall, while Ubiquitin was the least stable. The stability rankings of the seven candidate reference genes from the ΔCt analysis were largely consistent with those obtained from geNorm, NormFinder, and BestKeeper. However, there are some discrepancies in the rankings of the candidate reference genes.

RefFinder analysis

RefFinder is a comprehensive software tool that incorporates four analytical methods, GeNorm, NormFinder, BestKeeper, and ΔCt, to determine the overall ranking of candidate reference genes23. The results in Table 4 indicate significant variability in the stability of the seven candidate reference genes. Actin demonstrated the highest stability at 1 dai, 5 dai, and through all samples, while eIF5A3 emerged as the most stable at 3 dai. Combining the Ct values, GeNorm, NormFinder, BestKeeper, ΔCt, and RefFinder results, EF1α, eIF5A3, and Tubulin can be selected as reference genes for evaluating gene expression levels in different potato samples.

Validation of reference genes

To assess the reliability and stability of EF1α as a normalization factor for relative gene expression analysis, we performed RT-qPCR technology to investigate the expression levels of four key genes involved in the plant-pathogen interaction pathway: RBOHC (Soltu.Q9.Chr03_A20009820.g), WRKY24 (Soltu.Q9.Chr06_A2001677.g), MPK3 (Soltu.Q9.Chr06_A30016090.g), and CPK32 (Soltu.Q9.Chr10_A10025850.g). As illustrated in Fig. 5 indicate that when EF1α is used as the reference gene, the expression levels of these four genes show higher similarity to the log2 fold changes observed in RNA-seq data. In contrast, when Ubiquitin is used as the reference gene, the expression levels of these four genes vary greatly. Therefore, the results of the stability analysis are reliable. FPKM expression levels of the four genes are provided in Supplementary Table S4.

Fig. 5
Fig. 5The alternative text for this image may have been generated using AI.
Full size image

Gene expression with EF1α and Ubiquitin as reference genes for gene analysis: Q represents Qingshu 9 and L represents Leshu 1. 1, 3, and 5 indicate the 1, 3, and 5dai after inoculation with the pathogen P. atrosepticum, the units of RNA-Seq data are multiples of log2 Fold Change.

Discussion

Potato blackleg disease is a serious bacterial disease that has severely impacted potato production and storage for a long time, and potato resistance is one of the most effective and environment-friendly approaches used to manage the disease24,25. For example, expression profiling studies of resistance-related genes might be undertaken to get an insight into the resistance mechanisms in the potato- P. atrosepticum interaction at the molecular level26. The accuracy of such studies critically depends on the selection of appropriate internal reference genes for quantitative analysis. However, there has been limited research on this subject. Therefore, this study is of great significance for subsequent calculations of the expression levels of related disease-resistance genes in potatoes infected with P. atrosepticum.

RT-qPCR technology is essentially a relative quantification method, which evaluates the relative expression levels of target DNA molecules by comparing them to the expression level of reference genes27. However, various experimental and biological conditions influence the expression stability of reference genes. Research has shown that under specific conditions, different plants and pathological states can cause significant variations in the expression stability of reference genes. For example, in Haloxylon ammo dendron under various types of abiotic stress (such as heat, drought, salinity, ABA, and diurnal rhythms), different reference genes (ALB, RPII, H3) exhibited varying levels of expression stability. ALB was identified as the optimal reference gene under ABA treatment and diurnal rhythms, RPII exhibited the greatest stability under drought stress, and the H3 gene was the most stable under heat stress28. Similar phenomena have been confirmed in other experiments, such as Yao Ziting et al.29, who found in their screening of reference genes for Pitaya canker disease that GAPDH was the least stable. In contrast, Yao Quansheng30, in the study of mango infected with bacterial angular spot, discovered that GAPDH exhibited relatively good stability.

To assess the stability of reference genes, various statistical algorithms and software tools such as geNorm, NormFinder, BestKeeper, and RefFinder are commonly employed31,32,33. These tools are capable of evaluating the expression stability of multiple genes. Discrepancies in results among different analysis software can be attributed to the inherent variability in algorithms and each tool’s specific advantages and limitations. For instance, in a study by Dai Yuli et al.34 on maize northern leaf blight, NACA was identified as the most stable gene in geNorm, while TUBA was deemed most stable in NormFinder, and EF1α was the least stable in both geNorm and NormFinder. However, EF1α emerged as the most stable gene and Actin the least stable in BestKeeper. Similarly, in a study on sweet osmanthus (Osmanthus fragrans Lour)35 where geNorm and NormFinder produced consistent results, with OfUBC2 and OfACT as the most stable genes and Of18S as the least stable. Conversely, BestKeeper yielded different results, designating OfRAN1 as the most stable reference gene. Therefore, relying on a single data analysis tool may not offer a comprehensive and objective assessment of candidate reference genes. It is crucial to integrate results from multiple analytical tools to ensure the accuracy and reliability of RT-qPCR experiments.

The selection of appropriate reference genes is critical for accurate gene expression profiling in potatoes infected with P. atrosepticum. EF1α has demonstrated stable expression in multiple independent studies. For instance, during biotic stress (late blight) and abiotic stresses (cold and salt stress) in potatoes, EF1α demonstrates relatively stable expression, while the expression of other reference genes shows variations under these stress conditions36. Luo Jiajia37 screened for stably expressed reference genes in different tissues (roots, stems, leaves, flowers, and petioles) of Centella asiatica, while He Yongxiang38 evaluated the stability of reference genes in OidiumheveaeSteinm under high-temperature stress across distinct developmental stages (conidial stage, germination stage, infection stage, growth stage, and sporulation stage). Both studies demonstrated that EF1α exhibited superior stability, establishing as an ideal reference gene. In this study, the expression levels of four DEGs related to plant-pathogen interactions were analyzed to verify the applicability and reliability of EF1α as a reference gene under potato infection by P. atrosepticum. The results indicated that when Ubiquitin, which has lower stability, was used as the reference gene, there were significant differences in the relative expression levels of the DEGs. However, when EF1α was used as the reference gene, the relative expression levels of the DEGs aligned more closely with expectations based on previous RNA-seq data, further confirming the applicability and reliability of EF1α as a reference gene.

This study evaluated the expression stability of seven candidate reference genes using RT-qPCR across two potato cultivars inoculated with P. atrosepticum, with sampling conducted at three distinct time points post-inoculation. The results demonstrated that EF1α exhibited superior stability compared to other candidate genes. Consequently, EF1α was identified as a reliable reference gene for normalizing the relative expression levels of target genes in potato tissues following infection by P. atrosepticum.

Conclusion

This study analyzed the stability of seven reference genes (EF1α, Tubulin, eIF5A3, Actin, CYP3, GAPDH, Ubiquitin) using geNorm, NormFinder, BestKeeper, Delta CT, and RefFinder. geNorm analysis identified eIF5A3 and Actin as the most stable, followed by EF1α; Ubiquitin was least stable. Consistent results from NormFinder, BestKeeper, and Delta CT indicated that EF1α exhibited highly stable, with GAPDH and Ubiquitin less stable. Ultimately, RefFinder concluded EF1α was the most stable reference gene.