Selection and validation of reference genes for gene expression analysis in switchgrass (Panicum virgatum) using quantitative real-time RT-PCR

Abstract

Switchgrass (Panicum virgatum) has received a lot of attention as a forage and bioenergy crop during the past few years. Gene expression studies are in progress to improve new traits and develop new cultivars. Quantitative real time PCR (qRT-PCR) has emerged as an important technique to study gene expression analysis. For accurate and reliable results, normalization of data with reference genes is essential. In this work, we evaluate the stability of expression of genes to use as reference for qRT-PCR in the grass P. virgatum. Eleven candidate reference genes, including eEF-1α, UBQ6, ACT12, TUB6, eIF-4a, GAPDH, SAMDC, TUA6, CYP5, U2AF, and FTSH4, were validated for qRT-PCR normalization in different plant tissues and under different stress conditions. The expression stability of these genes was verified by the use of two distinct algorithms, geNorm and NormFinder. Differences were observed after comparison of the ranking of the candidate reference genes identified by both programs but eEF-1α, eIF-4a, CYP5 and U2AF are ranked as the most stable genes in the samples sets under study. Both programs discard the use of SAMDC and TUA6 for normalization. Validation of the reference genes proposed by geNorm and NormFinder were performed by normalization of transcript abundance of a group of target genes in different samples. Results show similar expression patterns when the best reference genes selected by both programs were used but differences were detected in the transcript abundance of the target genes. Based on the above research, we recommend the use of different statistical algorithms to identify the best reference genes for expression data normalization. The best genes selected in this study will help to improve the quality of gene expression data in a wide variety of samples in switchgrass.

Figure 1. Specificity of real-time RT-PCR amplification.

A) Melting curve of the 11 reference genes showing a single pick (each including three technical replicates of the cDNA pool of the total samples used in this study). B) Agarose gel (1.5%) showing amplification of a specific PCR product of the expected size for each gene tested in this study.

Figure 2. Cycle threshold (Ct) values of the candidate reference genes across the experimental samples.

Box-plot graph of Ct value shows the median values as lines across the box. Lower and upper boxes indicating the 25^th percentile to the 75^th percentile. Whiskers represent the maximum and minimum values.

Figure 3. Gene expression stability values (M) and pairwise variation (V) of the candidate reference genes calculated by geNorm.

A. Ranking of the gene expression stability performed in all the samples, abiotic stress samples, tissue samples and leaf and stem samples. The least stable genes are on the left and the most stable genes on the right. B. Pairwise variation (V_n/V_n+1) was analyzed between the normalization factors NFn and NFn+1. Asterisk indicates the optimal number of reference genes required for normalization.

Figure 4. Relative quantification of SGCesAs genes using different combinations of reference genes for normalization.

A. Gene expression normalized with the best combination of reference genes selected by NormFinder for all set of samples under study. B. Gene expression normalized with the best combination of reference genes selected by NormFinder for tissue samples. C. Gene expression normalized with the best combination of reference genes selected by NormFinder for samples under abiotic stress. D. Gene expression normalized with the most stable reference genes selected by geNorm for all samples under study. E. Gene expression normalized with the most stable reference genes selected by geNorm for tissue samples. F. Gene expression normalized with the most stable reference genes selected by geNorm for samples under abiotic stress. G. Gene expression normalized with the less stable genes identify by both algorithms for all samples under study. H. Gene expression normalized with the less stable genes identify by both algorithms for tissue samples. I. Gene expression normalized with the less stable genes identify by both algorithms for samples under abiotic stress.