Computational protocol: Diel pattern of circadian clock and storage protein gene expression in leaves and during seed filling in cowpea (Vigna unguiculata)

Similar protocols

Protocol publication

[…] We identified candidate reference genes from legumes using the gene expression atlas from Medicago truncatula [] and a set of genes found suitable for normalization in soybean []. We used the accession numbers to identify cowpea genomic sequences by BLAST (harvest-web.org). Scaffolds were retrieved and using legume translated mRNAs, we identified putative mRNAs from cowpea using Genewise [] and Noble VuGEA (Additional file : Table S1). The genes used were Β-ACTIN (ACT), ACTIN 2/7 (ACT27), CYCLOPHYLIN (CYP), ELONGATION FACTOR 1-A (EF1A), ELONGATION FACTOR 1-B (EF1B), ALPHA TUBULIN (TUA4), BETA TUBULIN (TUB4), ASK-INTERACTING PROTEIN 16 (SKIP16) and a HYPOTHETICAL UNKNOWN PROTEIN from soybean (UKN2). The genes related to protein storage accumulation were LEGUMIN (LEG), LEGUMINJ (LEGJ) and COVICILIN (CVC). Circadian clock related genes were GIGANTEA (GI), TIMING OF CAB EXPRESSION1 (TOC1), LATE ELONGATED HYPOCOTYL (LHY), and EARLY FLOWERING 3(ELF3). Primers were designed using the software PCRefficiency (http://srvgen.upct.es/efficiency.html) as described previously [] (Additional file : Table S1). Primers were tested for stable, single and clear amplification products by end-point PCR with genomic DNA, visualized on 1.5% agarose gels (Additional file : Figure S1) and by quantitative PCR to assess the melting profile of the PCR products (Additional file : Figure S2). [...] For the identification of stable reference genes during different developmental stages and tissue, PCR efficiencies and CT values were used in a web-pipeline that contains the different PCR analysis softwares Bestkeeper, Normfinder, Delta CT and Genorm. PCR efficiency was calculated as described before []. Data from different analysis was pooled and ranked using Rank-Aggreg (Pihur and Datta, []). We used the software Geomean to obtain a ranking value of the candidate reference genes [].Statistical analysis of diurnal gene expression profiles for clock relates genes and storage protein related genes was performed using the normalized cycle threshold (Ct) values calculated as described previously []. A PCR efficiency of 2 for all primer combinations was used for the calculation of normalized expression (NE) based on efficiency calculations, which were performed as described previously with the qpcR R package [, ]. Average efficiencies were 1.98 for VuEF1A, 1.99 for VunGI, 1.97 for VunELF3, 1.95 for VunTOC1, 1.99 for VunLHY, 1.99 for VunCVC, 1.93 for VunLEG and 1.99 for VunLEGJ. JTK-Cycle method was applied for the determination of existence of a circadian biological rhythm represented in the transcriptome data [] using the R package “MetaCycle” that provides functions and methods (JTK_CYCLE, Lomb-Scargle and ARSER) for detecting rhythmic signals from time series datasets (https://cran.r-project.org/web/packages/MetaCycle/index.html). JTK_CYCLE results include the P value (Pval, significative if P < 0.05), period (Per), phase (Phase) and amplitude (Amp). Period is defined as the time between two consecutive peaks. Phase is considered as the time point with the peak and amplitude is the difference between the peak (or minimum) and the mean value of the wave.Statistical analysis for gene expression was performed using group-wise comparison with the REST program []. Phenotypic data were analysed for homogeneity of variance with the Fligner-Killeen test in R. The parameters showing homogeneity of variance were analysed using ANOVA and Tukey’s HSD test, while the non-parametric data were analysed using Wilcoxon signed rank test with continuity correction in R version 3.2.3. […]

Pipeline specifications