Selection and validation of reference genes for RT-qPCR normalization in different tissues of milk thistle (Silybum marianum, Gaert.)

Highlights

• S. marianum is an officinal plant important for silymarin and for other productions.

• Ten reference genes in six tissues were evaluated for gene expression studies.

• MON1 is the most stable gene in all datasets according to RefFinder.

• In the fruits MON1, FAD2-2, ACT and NDHF are the top ranked reference genes.

• Analysis of CHS1, PRX1 and ACP1 confirmed the importance accurate RGs selection.

Abstract

Quantitative reverse transcription PCR is a sensitive technique for evaluating transcriptional profiles in different experimental datasets. To obtain a reliable quantification of the transcripts level, data normalization with stable reference genes is required. Stable reference genes are identified after analysis of their transcripts profile in every new experiment and species of interest. In Silybum marianum, a widely cultivated officinal plant, only few gene expression studies exist, and reference genes for RT-qPCR studies in the diverse plant tissues have never been investigated before. In this work, the expression stability of 10 candidate reference genes was evaluated in leaves, roots, stems and fruits of S. marianum grown under physiological environmental condition. The stability values for each candidate reference gene were calculated by four canonical statistical algorithms GeNorm, NormFinder, Bestkeeper and ΔCt method in different subsets of samples, then they were ranked with RefFinder from the most to the least suitable for normalization.

Best combinations of reference genes are finally proposed for different experimental data sets, including all tissues, vegetative, and reproductive tissues separately. Three target genes putatively involved in important biosynthetic pathway leading to key metabolites in the fruits of milk thistle, such as silymarin and fatty acids, were analyzed with the chosen panels of reference genes, in comparison to the ones used in previous papers.

To the best of our knowledge, this is the first report on a reliable and systematic identification and validation of the reference genes for RT-qPCR normalization to study gene expression in S. marianum.

Introduction

Milk thistle (Silybum marianum Gaertn.) is an important officinal plant worldwide. It has been used as a medicinal herb for more than 2000 years (Morazzoni and Bombardelli, 1995) and recently studies addressing the use of S. marianum in the treatment of several pathologies have increased (Camini and Costa, 2020). Despite that, the species has not been completely domesticated (Martinelli, 2019). Its health-promoting properties are attributable to the presence in the fruit integument of a group of flavonolignans called silymarin, which has been widely investigated for the treatment of human liver diseases (Abenavoli et al., 2011) and for many other effects such as cancer prevention, antioxidant and neuroprotective activities (Kren and Walterova, 2005, Lama et al., 2019, Lani et al., 2015, Li et al., 2010). Due to its characteristics, milk thistle is one of the 10 bestselling herbal products in the US (Smith et al., 2017) and silymarin has been used as bioactive component of drugs for almost 50 years (Albrecht et al., 1992).

Apart from silymarin, milk thistle fruits (cypselae – fruits containing a single seed) have a high oil content (20–30%) (Martinelli et al., 2016). Prior silymarin extraction it is necessary to remove oil from the fruits, making oil an important byproduct of silymarin production. Moreover, its high fruit and biomass yield under low input conditions makes this species an interesting multipurpose crop for marginal Mediterranean environments (Afshar et al., 2014, Domínguez et al., 2017, Sulas et al., 2008).

Only in 2019, and limited to the PubMed database, more than 300 studies have been published dealing with S. marianum medicinal properties, suggesting that milk thistle is a species of interest in the current scientific scenario worldwide. Despite that, as for most non-model crop, only limited molecular studies exist and specific analytical protocols for functional genomic and transcriptomic studies are not available yet. In these species, very few genes involved in specific biosynthetic pathways have already been identified and their transcriptional patterns studied. However, despite next generation sequencing technology has enabled a quick and cost-effective analysis of whole genomes and transcriptomes, S. marianum genomic information is still limited and only a representative first draft whole genome sequence (Accession number: ASM154182v1) and transcriptome (Accession number: PRJNA298444) has been recently released (Lv et al., 2017).

The discovery of genes involved in silymarin biosynthesis and accumulation in fruits integument during development (El-Garhy et al., 2016, Elsayed et al., 2019, Lv et al., 2017, Sanjari et al., 2015, Soliman et al., 2018, Torres and Corchete, 2016) or in the synthesis of other valuable constituents of the fruits (Alemardan et al., 2013) are interesting research topics in S. marianum. The identification of key genes involved in fruit shattering is another milestone toward the genetic improvement of this crop (Martinelli, 2019). Moreover, studies regarding leaves variegation recently provided new information in this species about physiological adaptation to heat, UV radiation or defence from herbivores but, again, the involvement of specific genes is not known yet (Shelef et al., 2019). Finally, molecular studies of roots would be useful for a better comprehension of this tissue in order to improve the silymarin production by hairy root cultures (Hasanloo et al., 2008, Rahnama et al., 2013).

Quantitative real-time PCR (RT-qPCR) is an efficient strategy to estimate transcripts expression levels and the accuracy of its results is largely due to the expression stability of the reference genes. Accurate normalization of transcriptional data using selected reference controls enables the comparison of mRNA concentrations across samples (Bustin et al., 2009). Many studies have reported significant variations in the expression of reference genes in different experimental conditions and plant tissues. Therefore, the identification of reference genes with stable expression across samples, and their validation in every new experiment, cell type, tissue, species is a pre-requisite for a reliable transcriptional analysis (Barbierato et al., 2017). As reported in the Minimum Information for Publication of Quantitative Real-Time PCR Experiments (MIQE) guide, normalization using a single reference gene is no more acceptable and “the optimal number and choice of reference genes must be experimentally determined, and the method reported” (Bustin et al., 2009).

Genes involved in the maintenance of basic cellular functions, named housekeeping genes, are good candidates to be used for normalization as they are expected to be stably expressed in the cells (Wu et al., 2016). Tubulin alpha (TUBα), Tap42-interacting protein of 41 kDa (TIP41), NADH dehydrogenase (NADH), 18S rRNA (18S), Actin (ACT), ubiquitin conjugating enzyme E2 (UBC), monensin sensivity1 (MON1) and protein phosphatase 2A (PP2A) genes are among the most common reference genes in plants (Bustin, 2002, Czechowski et al., 2005, Expósito-Rodríguez et al., 2008, Goidin et al., 2001, Kim et al., 2003). Nevertheless, it has been reported that these genes are expressed diversely according to genotypes, tissues, stress conditions or developmental stages (Dheda et al., 2004, Garrido et al., 2020, Gutierrez et al., 2008, Nikalje et al., 2018, Reddy et al., 2013, Sinha et al., 2015). Therefore, the assessment of their expression reliability under certain conditions and between different tissues is highly recommended to ensure the accuracy of results.

In this study, 10 reference genes have been analyzed with the aim to provide an appropriate selection of reference genes for RT-qPCR data normalization in S. marianum. The stability of the selected candidates was assessed in different physiological tissues that can be site of expression of interesting target candidate genes, such as fruits at different developmental stages but also leaves, stems and roots. Moreover, the selected reference genes were validated by analyzing the transcriptional profiles of target genes involved in silymarin and fatty acids biosynthesis. To the best of our knowledge, the current study is the first report on a systematic analysis for selection of reference genes in various milk thistle tissues and development stages and will boost genomic and functional studies in this species.