Selection and Validation of Reference Genes for RT-qPCR Analysis in Spinacia oleracea under Abiotic Stress

Abstract

Reverse transcription quantitative real-time polymerase chain reaction (RT-qPCR) is an accurate and convenient method for mRNA quantification. Selection of optimal reference gene(s) is an important step in RT-qPCR experiments. However, the stability of housekeeping genes in spinach (Spinacia oleracea) under various abiotic stresses is unclear. Evaluating the stability of candidate genes and determining the optimal gene(s) for normalization of gene expression in spinach are necessary to investigate the gene expression patterns during development and stress response. In this study, ten housekeeping genes, 18S ribosomal RNA (18S rRNA), actin, ADP ribosylation factor (ARF), cytochrome c oxidase subunit 5C (COX), cyclophilin (CYP), elongation factor 1-alpha (EF1α), glyceraldehyde-3-phosphate dehydrogenase (GAPDH), histone H3 (H3), 50S ribosomal protein L2 (RPL2), and tubulin alpha chain (TUBα) from spinach, were selected as candidates in roots, stems, leaves, flowers, and seedlings in response to high temperature, CdCl₂, NaCl, NaHCO₃, and Na₂CO₃ stresses. The expression of these genes was quantified by RT-qPCR and evaluated by NormFinder, BestKeeper, and geNorm. 18S rRNA, actin, ARF, COX, CYP, EF1α, GAPDH, H3, and RPL2 were detected as optimal reference genes for gene expression analysis of different organs and stress responses. The results were further confirmed by the expression pattern normalized with different reference genes of two heat-responsive genes. Here, we optimized the detection method of the gene expression pattern in spinach. Our results provide the optimal candidate reference genes which were crucial for RT-qPCR analysis.

Figure 1

Specificity of primers for ten candidate reference genes in PCR and their amplicon sizes. The names of ten candidate reference genes including 18S ribosomal RNA (18S rRNA), actin, ADP ribosylation factor (ARF), cytochrome c oxidase (COX), cyclophilin (CYP), elongation factor 1-alpha (EF1α), glyceraldehyde-3-phosphate dehydrogenase (GAPDH), histone H3 (H3), 50S ribosomal protein L2 (RPL2), and tubulin alpha chain (TUBα) are noted on each lane. M represents 100 DNA ladder.

Figure 2

Cq values of ten candidate reference genes in spinach detected by RT-qPCR. The medians of Cq values for these genes are represented by the line in the boxes, and the upper and lower quartiles of Cq values are represented by the upper and lower boundaries of the boxes. The whiskers represent the ranges for the bottom 25% and the top 25% of these Cq values. Small circles represent outliers of these Cq values: (a) Cq values in organs including leaves, stems, roots, flowers, and seedlings, (b) Cq values in the leaves and roots under heat stress, (c) Cq values in the leaves and roots under heavy metal stress, (d) Cq values in the leaves and roots under NaCl, (e) Cq values in the leaves and roots under NaHCO₃, and (f) Cq values in the leaves and roots under Na₂CO₃. Ten candidate reference genes refer to Figure 1. Three biological replicates were taken for different organs/seedlings, as well as leaves and roots at each time point of all stress treatments.

Figure 3

Average expression stability values (M2) in different data sets obtained from the software geNorm. The stability values (M2) and stability rank were obtained from geNorm. A lower M2 value suggests higher stability: (a) stability rank obtained in organs including leaves, stems, roots, flowers, and seedlings, (b) stability rank obtained in the leaves and roots under heat stress, (c) stability rank obtained in the leaves and roots under heavy metal stress, (d) stability rank obtained in the leaves and roots under NaCl, (e) stability rank obtained in the leaves and roots under NaHCO₃, and (f) stability rank obtained in the leaves and roots under Na₂CO₃. Ten candidate reference genes refer to Figure 1. Three biological replicates were taken for different organs/seedlings, as well as leaves and roots at each time point of all stress treatments.

Figure 4

Pairwise variation (V) analysis of the candidate reference genes. The geNorm software was used to analyze the pairwise variation (V_n/V_n+1) between the normalization factors (NF) NF_n and NF_n+1 in order to determine the optimal number of candidate reference genes required for RT-qPCR data normalization. If V_n/V_n+1 < 0.15 (gray dotted line), the minimum value of n is the optimal number of genes required.

Figure 5

Expression pattern of SobZIP9 and SoHSFB2b normalized by different genes in response to heat stress. Expression patterns of two heat-induced genes, basic region-leucine zipper 9 (bZIP9, a) and heat stress transcription factor 2 b (HSFB2b, b), were normalized with actin, ADP ribosylation factor (ARF), and tubulin alpha chain (TUBα). Three biological replicates were taken for leaves at each time point of heat treatments. Data represent average ± SD, and ∗ indicates significant difference (p < 0.05).