Stability of endogenous gene expression
To identify the most suitable endogenous control genes, the level of transcript accumulation of the samples was verified with respect to two stress types, biotic (P. psidii) and abiotic (ASM), and different genotypes (C0 and VR clones). The expression profile of the 13 genes is shown in Figure 1 and Additional file 2. Genes 30S, 60S, EgIDH, TUB and UBQ had the highest expression levels (lowest Ct value) in both clones (Figure 1, Additional file 3). Except for PUBQ, the remaining genes did not have substantial differences in Ct value between clones. However, Ct variations due to ASM application and P. psidii inoculation were observed at different levels in several genes (Figure 1, Additional file 2).
To choose the best candidate endogenous genes, i.e., the most stably expressed genes following the pathogen and resistance inducer treatments, GeNorm [7] and GenEx version 4.3.6. software [13] were used. Vandesompele et al. [7] defines an expression stability mean value (M-value) as a parameter for quantification of the stability of candidate endogenous genes, in which a low M-value indicates more stable expression, making a certain candidate more appropriate for an endogenous gene control. This value is based on the geometric mean of multiple endogenous genes and the mean variation of a gene relative to all others in a sample group, according to the principle that the relationship between the expression of two ideal endogenous control genes is identical in all samples, and is independently of the experimental conditions.
In the present work, stability of expression was observed in several situations. Considering only C0 and VR tissue samples that did not receive any stimuli, eEF2 and EgIDH had the lowest mean values of expression stability (M-value: 0.157), i.e., these genes had more stable expression relative to the other genes evaluated (Figure 2a). Comparison of P. psidii-inoculated and uninoculated C0 and VR samples indicated that eEF2 and UBQ were the most stably expressed genes (M-value: 0.276) (Figure 2b). In ASM-treated and untreated leaves of both clones, CYP and elF4B had an M-value of 0.275 and were the most stably expressed genes (Figure 2c). Comparing ASM-treated, P. psidii-inoculated, and ASM-treated plus P. psidii-inoculated leaves of both clones, as well as their respective controls, EgIDH and UBQ genes had the lowest stability mean values (M-value: 0.321) (Figure 2d). In all cases, PUBQ was the least stably expressed gene because it had the highest M-values.
Chitinase gene relative expression
To confirm the stability of candidate endogenous gene expression according to geNorm software and to assure variation in gene expression in response to the experimental conditions (ASM and P. psidii treatments), the level of chitinase gene expression was assessed. This gene has been highly expressed in studies related to plant resistance to phytopathogens [14, 15] and in response to ASM application [14, 1, 16]. Chitinase gene expression levels were determined according to the number of amplification cycles needed to reach a fixed threshold (Ct) in the exponential phase of PCR and analyzed using the GenEx version 4.3.6 software. The data were normalized according to the most stable candidate endogenous genes (eEF2+EgIDH; eEF2+UBQ; and EgIDH+UBQ) identified by geNorm software under the various experimental conditions (Figure 2); genes EgIDH, 30S and PUBQ, which have been greatly employed as the sole normalizer genes in several studies, were also used. After data normalization with respect to endogenous genes, chitinase gene expression differences were verified among ASM-treated, P. psidii-inoculated, ASM-treated plus P. psidii-inoculated C0 (Figure 3a) and VR clones, and their respective controls.
The EgIDH gene has been employed in several studies as an endogenous reference gene for data normalization [17–19]; however, when this gene was used as the only normalizer for chitinase gene data, the results did not indicate changes in the relative expression of the chitinase gene under any experimental conditions (Figure 3a and 3b). Chitinase gene expression could be differentiated in ASM-treated C0 and VR clones (upregulated 5.58-fold and 3.32-fold, respectively), and ASM-treated plus P. psidii-inoculated C0 and VR clones (upregulated 7.70-fold and 4.82-fold, respectively), relative to the respective controls (Additional file 3); however, no difference was observed in P. psidii-inoculated C0 and VR clones (upregulated 1.48-fold and 1.87-fold, respectively).
The 30S gene has also been used in many studies as the only endogenous gene control [3]. When this gene was employed as the only normalizer for chitinase gene relative expression data, increased expression could only be detected in the C0 clone under for P. psidii-inoculated (upregulated 3.30-fold), ASM-treated (upregulated 3.81-fold), and ASM-treated plus P. psidii-inoculated (upregulated 3.54-fold) leaves (Additional file 3). In the VR clone, only ASM-treated leaves had an increase (upregulated 2.25-fold) in chitinase gene expression (Additional file 3).
The PUBQ gene had the highest variation in expression variations across experimental conditions, according to geNorm results (Figure 2). When used for chitinase gene data normalization, it was not useful in detecting changes in level of expression under the different experimental conditions (Figure 3). Only the VR clone had increased expression for P. psidii-inoculated (upregulated 3.13-fold), ASM-treated (upregulated 2.96-fold), and ASM-treated plus P. psidii-inoculated (upregulated 4.89-fold) leaves, compared with the respective controls (Additional file 3). For the C0 clone, an increase in expression relative to control was only observed in ASM-treated plus P. psidii-inoculated leaves (upregulated 2.01-fold).
Comparing tissue samples from C0 and VR clones that did not receive any stimulus, eEF2 and EgIDH were the most stably expressed genes according to geNorm (Figure 2a). When such genes were used together for chitinase gene data normalization, there was a substantial difference: chitinase expression was higher in the C0 than the VR clone (Figures 3a and 3b). After stimulus, the VR clone had chitinase gene expression increases higher for P. psidii-inoculated (upregulated 4.27-fold), ASM-treated (upregulated 5.58-fold), and ASM-treated plus P. psidii-inoculated (upregulated 7.70-fold) leaves. However, the C0 clone showed expression differences only for ASM-treated plus P. psidii-inoculated (upregulated 2.25-fold) leaves (Additional file 3).
eEF2 and UBQ were considered the most stably expressed genes in comparisons of expression levels of P. psidii-inoculated and uninoculated C0 and VR leaves. When used together for chitinase gene normalization, these genes helped to detect expression differences in both clones with respect to all variables relative to control (Figure 3 and Additional file 3). Cyc and Euk4 were the most stably expressed genes in the comparison between the expression levels of ASM-treated and untreated leaves. When used together for chitinase gene normalization, these genes only had changes in expression in ASM-treated leaves. The C0 clone had 2.75-fold for ASM-treated leaves and 4.02-fold higher expression for leaves treated both with ASM and inoculated with P. psidii relative to controls. VR clone had a similar behavior, upregulated 2.68-fold for ASM-treated leaves and 3.78-fold for ASM-treated plus P. psidii-inoculated leaves, relative to controls.
ASM-treated, P. psidii-inoculated, and ASM-treated plus P. psidii-inoculated clones, EgIDH and UBQ were the most stably expressed genes (Figure 2d). When used together for chitinase gene normalization, these genes best detected changes in chitinase expression with respect to all variables. In examining the effect of P. psidii inoculation, the C0 clone had a 7.33-fold and VR clone had a 4.79-fold increase relative to controls. For the ASM treatment effect, the C0 clone had an 8.41-fold increase in chitinase expression level and the VR clone a 9.45-fold increase. ASM treatment followed by P. psidii inoculation 5 days later led to a 12.28-fold increase in expression in the C0 clone and an 11.24-fold increase in the VR clone.