3.1 Verification of PCR product specificity and determination of CT value validity
The specificity of PCR products was detected by gel electrophoresis. It could be observed that each PCR product revealed a single band in agarose gel (Fig. S1). Besides, the melting curve of each gene showed a single peak (Fig. S2), which indicated that the primers of reference genes had strong specificity. According to the amplification efficiency (En) (ranged from 91.81% (EF1) to 103.04% (60s18)) and correlation coefficient (R2) (R² > 0.99), the results complied with the standard of qRT-PCR (Table 1).
Under various abiotic stresses, the expression levels of selected reference genes diversified greatly across all samples of C. auriculatum. As shown in Fig. 1, the ultimate value of CT was 31.66, and the minimum value was 13.18. The expression abundance of EF1 was the highest, and the maximum, minimum, and median values of CT were 38.64, 24.04, and 31.88, respectively. The maximum, minimum, and median CT values of EF1 expression were 21.10, 17.91, and 20.29, respectively. In addition, the reference genes showed significant variable expression. Compared with the other genes, SAMDC and TUB-α had a relatively narrow CT value range, indicating that they were more stable in term. These results also showed that for C. auriculatum, it is essential to select corresponding reference genes under certain conditions.
3.2 GeNorm analysis
GeNorm evaluates gene stability by calculating the M value. M value less than 1.5 usually represents the stable expression of the gene. M values and the ranking list of the 12 selected genes under various stresses were provided by geNorm (Fig. 2), and M values across all samples were lower than 1.5. Under heat stress, the expression of SAMDC, CYP, and ACT7 was the most stable (Fig. 2A). Under cold stress, the face of ACT7, MUB3, and SAMDC was the most stable (Fig. 2B). Under drought stress, TUB-α, GAPDH, and EF2 were observed as the most stably expressed (Fig. 2C). Under waterlogging stress, TUB-α, SAMDC, and 60s13 were most stably expressed (Fig. 2D). Under salt stress, the expression of SAMDC, ACT7, and TUB-α was the most stable (Fig. 2E). In conclusion, ACT7 showed high stability under heat, cold, and salt stresses. Under both drought and waterlogging stresses, the expression of TUB-α was relatively more stable. SAMDC had the highest stability under heat, cold, waterlogging, and salt stresses. MUB3 was stably expressed only under cold stress. Besides, in all samples, ACT7, TUB-α, and GAPDH were identified as stably expressed genes (Fig. 2F).
To find out the appropriate number of selected genes, geNorm software introduced paired variation analysis to calculate the Vn/n+1 value 38. In this study, the Vn/n +1 was less than 0.15 under heat, cold, drought, waterlogging, and salt stress, which indicated that the n of gene combinations should be 2. In all samples, the paired variation was 0.165 for V2/3, and 0.136 for V3/4 (Fig. 3), indicating that adding a third gene would obviously affect the analysis result. Therefore, the best reference gene combination in all samples of C. auriculatum was ACT7, TUB-α, and GAPDH.
3.3 NormFinder analysis
To better analyze the stability of the 12 selected genes from C. auriculatum, NormFinder software was applied for further study. The data showed that, under heat stress, the most stable genes were EF2 (0.243), TUB-β (0.243), and SAMDC (0.359) (Fig. 4A). Under cold stress, the expression of TUB-β (0.107), CYP (0.355), and GAPDH was more stable (0.469) (Fig. 4B). Under drought stress, the expression of GAPDH (0.165), TUB-α (0.217), and EF2 (0.290) was more stable (Fig. 4C). The top three genes stably expressed were SAMDC (0.084), TUB-α (0.139), and CYP (0.395) under waterlogging stress, and EF2 (0.284), Elf (0.307), and 60s13 (0.499) under salt stress (Fig. 4D-E). Besides, among all samples, EF2 (0.048), GAPDH (0.048), and CYP (0.258) were the most stable (Fig. 4F). In addition, the stability ranking lists provided by NormFinder and geNorm were inconsistent.
3.4 BestKeeper analysis
Based on CV and SD values, BestKeeper ranked the expression stabilities of the 12 selected genes from C. auriculatum, and the ranking list was shown in Table 2. Under heat stress, MUB3 (0.64 ± 0.09), TUB-α (1.09 ± 0.30), and ACT7 (3.06 ± 0.62) expressed more stable than other genes. Under cold stress, the expression ACT7 (1.89 ± 0.39), MUB3 (1.99 ± 0.28), and SAMDC (2.95 ± 0.61) was more stable. The expression of MUB3 (2.31 ± 0.33), ACT7 (3.16 ± 0.63), and SAMDC (4.43 ± 0.85) were the most stable under drought stress. Under waterlogging stress, the most stable genes included MUB3 (0.46 ± 0.07), ACT7 (0.93 ± 0.19), and TUB-α (1.73 ± 0.46). Under salt stress, MUB3 (0.47 ± 0.07), TUB-α (0.99 ± 0.27), and SAMDC (2.01 ± 0.44) expressed more stably. Meanwhile, in all samples, MUB3 (0.72 ± 0.10), ACT7 (1.57 ± 0.32), and TUB-α (1.62 ± 0.44) were most stably expressed.
3.5 RefFinder analysis
Three types of software had been applied in the stability evaluation of the 12 selected genes. However, because of the algorithm-specific preference, the software produced different results. Therefore, the software RefFinder, which integrates the above results, was applied to make an overall ranking of the 12 selected genes among all samples. According to the ranking result produced by RefFinder, the expression of EF2, CYP, TUB-α, SAMDC, and EF2 was more stable under heat, cold, drought, waterlogging, and salt stress, respectively. Meanwhile, EF2 was listed as the optimal gene among all candidate genes. Besides, except for drought and salt stresses, 60s18 was the most unstable gene (Table 3).
3.6 Stability verification of selected genes associated with CaWRKY1
To validate the stabilities of potential genes obtained by comprehensive analysis, the top two and the bottom two genes in the ranking list (Table 3) were picked out to quantify the expression of CaWRKY1. It could be obtained that, under heat stress, compared with the top two genes and their combination, the relative expression level of CaWRKY1 reached a peak (5.23,5.31, and 5.25) at 12h. At the same time, compared with the bottom genes, the relative expression level of CaWRKY1 increased or decreased differently (Fig. 5A). Under cold, drought, waterlogging, salt stresses, and different stresses (Fig. 5B-F), compared with the top two genes and their combinations, the changing trends of CaWRKY1 relative expression level were similar. In addition, compared with the bottom genes of the ranking table, the expression-changing trends of CaWRKY1 were significantly different from the former. Therefore, these results indicated that the data obtained from the RefFinder algorithm was accurate and reliable.