Validation of Stable Housekeeping Genes for Quantitative Real Time PCR in Golden Syrian Hamster

RT-qPCR is a sensitive technique that provides accurate and repeatable quantification of nucleic acids through the incorporation of fluorescent molecules into genetic material during amplification with real-time analysis of the increase in fluorescence (Nolan et al., 2006). The technique is commonly used to validate transcriptomic data and to determine relative levels of gene expression. To determine changes in gene expression, RT-qPCR requires a comparison between the target gene and one or multiple housekeeping genes that maintain consistent levels of expression during physiological challenges. Generally, housekeeping genes are selected based on their involvement in structural mechanisms or fundamental metabolic processes, such that constant expression can be assumed. Bestkeeper, geNorm and NormFinder are software programs commonly used to evaluate the stability of housekeeping genes. However, the most frequently used housekeeping genes are not always stably expressed in different tissues or different physiological conditions (Bages et al., 2015). Gapdh and b-actin have been proposed as universal standards for RT-qPCR studies, but the expression levels of these genes can still vary in different tissues. Moreover, a housekeeping gene with stable expression in one organism may not be suitable for normalization of gene expression in another organism (Kanakachari et al., 2016). 
       
The golden Syrian hamster (Mesocricetus auratus) is the model-of-choice for studies of many infectious diseases (Miao et al., 2019), cardiovascular disease (Dillard et al., 2010) and many cancers (Li et al., 2018). Currently, there is a general lack of hamster-specific commercial reagents for evaluating protein expression which limits the use of hamsters as models in these studies due to the inability to evaluate transcriptional changes during disease. A previous study lists several primer-probe sets that can be used in RT-qPCR assays (Zarubaev et al., 2011); however, this list is severely limited in comparison to those available to humans and mice in Primer Bank databases (https://pga.mgh.harvard.edu/primerbank/) (Wang et al., 2012).  In the present study, we evaluate eight housekeeping genes for RT-qPCR in lung, liver, spleen and pancreatic tissue of GSHs and make recommendations for their appropriate use in future studies.

Ethics statement and animals
 
The study was carried out during the period of June 2019 to October 2019, in the Animal Experiment Center, Henan University of Chinese Medicine, Zhengzhou. It‘s approved by the animal ethics committee of Henan University of Chinese Medicine and performed in accordance with established International Guiding Principles for Animal Research. Forty golden Syrian hamsters (5-6 weeks old) were purchased from Vital River Laboratory Animal Technology (Beijing, China) and maintained in the animal care facility of Henan University of Chinese Medicine. Ten individuals were randomly chosen and sacrificed by cervical dislocation. Samples of lung, liver, spleen, pancreas were collected in TRIzol Reagent (Life Technologies, CA, USA) and frozen in liquid nitrogen and stored at -80^oC until analysis.
 
RNA isolation and cDNA synthesis
 
Approximately 50 mg of each tissue was used for total RNA extraction with TRIzol Reagent. Quantity and quality of RNA were determined at 260 and 280 nm by NanoDrop Spectrophotometer (Thermo Fisher Scientific, MA, USA). Integrity of RNA was verified by staining of 18s and 28s rRNA band on a 1% agarose gel. RNA (2μg) from each sample was treated with DNase I to remove genomic DNA. Reverse transcription was performed using the Prime Script RT Reagent kit (Takara, Japan). Reactions were performed in an Applied Biosystems thermocycler with the following settings: 37^oC for 15 min and 85^oC for 5 min, followed by 42^oC for 2 min, then cooled to 4^oC. cDNA samples were stored at -20^oC until RT-qPCR analysis was performed.
 
Primer design
 
There is currently, no standardized method for identifying ideal candidates for housekeeping genes. Eight candidate housekeeping genes selected based on studies in humans and mice, which showed stable expression in different physiological conditions (Eisenberg and Levanon 2013, Eissa et al., 2017). These genes were Gapdh, Actb, B2m, Hprt1, Tubb, Rpl13a, Rps18 and Ywhaz. The sequences were chosen from Gene bank and primers were designed by Primer-Blast (https://www.ncbi.nlm.nih.gov/tools/primer-blast/) (Ye et al., 2012) (Table 1).
 

 
Quantitative RT-PCR
 
RT-qPCR reaction was performed in a StepOnePlus Real-Time PCR System using the TB Green Premix Ex Taq II (Takara, Japan) following the manufacturer’s instructions. The PCR reaction consisted of 10μL of TB Green Premix Ex Taq II, 0.8μL of forward and reverse primers (10μM), 0.4μL of ROX Reference Dye, 2μL of template cDNA in a total volume of 20μL. Cycling was performed using the StepOne Software v2.3, with the cycles set at 15 sec at 95^oC, 1 minute at 60^oC and 15 sec at 95^oC, followed by 40 rounds of 5 sec at 95^oC and 30 sec at 60^oC. To verify the specificity of each primer, we also performed a melting-curve analysis (60-95^oC with fluorescence measured every 0.5^oC). All reactions were performed twice to ensure technical reproducibly of the assays. The average standard deviation within duplicates studied was 0.5 cycles.
 
Primer’s efficiency
 
The minimum information for publication of quantitative RT-PCR experiments (MIQE) guidelines (Bustin et al., 2009) recommends analysis of RT-qPCR efficiencies for each primer pair using standard curves (5-point, 10-fold serial dilution of pooled cDNA including equal amounts from the sample sets). The mean cycle threshold Ct) values for each serial dilution were plotted against the logarithm of the cDNA dilution factor and calculated according to the equation.
 
Where,
In slope is the gradient of the linear regression line (Eissa et al., 2017). The linear dynamic range was determined by the standard and correlation coefficients (R²) for each gene as reported.
 
Evaluation of the stability of candidate reference genes
 
To evaluate the stability of the reference genes expressed as Ct values, we used StepOne Software v2.3, which applies the same statistical algorithm as geNorm. The comparative analysis of the stability of the housekeeping genes was performed in four software programs: Best keeper, geNorm, Normfinder and Ref Finder. Best keeper calculates the stability of candidate gene expression based on the Ct value of each candidate gene (Pfaffl et al., 2004). geNorm defines and ranks the reference gene based on an M value, where in the M value is defined as the average pairwise variation of the gene of interest relative to all other control genes (Vandesompele et al., 2002). Normfinder simultaneously evaluates intra- and inter-group variation and then combines the two to produce stability values. Stability values indicate the amount of systematic error introduced when studying genes (Andersen et al., 2004). The comparative delta Ct method is used to assess the most stable housekeeping genes by comparing relative expression within each tissue sample (Silver et al., 2006). Data from each of the four algorithms were then combined in RefFinder software (a web comprehensive evaluation platformhttps:https://www.heartcure.com.au/reffinder/)  which integrates the data from each of these analyses to rank the overall stability of each gene based on the geometric means (Xie et al., 2012). All data are expressed as means SD. GraphPad Prism 8.0 (La Jolla, CA, USA) was used for statistical procedures and graph plotting.

Primer specificity and efficiency
 
We assessed primer specificity and efficiency for each pair of primers. The dissociation curve confirmed the amplicon specificity. A single peak in the melting curve analyses for eight primers pairs indicated high specificity. The amplification efficiency for all selected primers ranged from 90% to 110%, the R² values were greater than 0.99 (Table 2).
 

 
Housekeeping gene expression profiles
 
The expression levels of the 8 housekeeping genes were evaluated using Ct and descriptive statistics (mean, SD, median, min, max) in all tested samples (Table 3). The mean Ct values for housekeeping genes ranged from 16 to 33, with most between 18 and 25. Ct values between 16 and 32 are considered normal for housekeeping genes, thus in almost all cases, the genes we investigated were found to be reasonable candidates. Two exceptions were identified, including Hprt1 and Tubb expression in pancreatic tissue (Ct values 33.23 and 32.77 respectively), indicating relatively low expression.
 

 
Appropriateness of housekeeping genes
 
To evaluate and rank the suitability of the housekeeping genes in lung, liver, spleen, pancreas, we applied four different algorithms to evaluate the stability of gene expression: NormFinder, Bestkeeper, geNorm, delta Ct method (Table 4). These data were then integrated in RefFinder to identify the best candidates for housekeeping genes. Fig 1 illustrates the comprehensive ranking of gene stability in lung, spleen, liver, pancreas. Gene stability varied among tissues. In lung tissue, the most stable gene was Actb and the least stable was Ywhaz (Fig 1A). In the liver, the most stable gene was Hprt1 and the least stable was Gapdh (Fig 1B). In splenic tissue, the most stable gene was Rpl13a and the least stable was Gapdh (Fig 1C). The most stable genes were Tubb and the least stable gene was Rps18 in pancreas (Fig 1D).
 

 

       
In these studies, using appropriate housekeeping gene is critical for generating reference points upon which to compare the expression levels of other genes.  (Bahr et al., 2009) employed microarray technology to identify species-specific housekeeping genes for Chinese Hamster Ovary (CHO) cells and provided a critical expression platform for gene expression analysis. However, to date, few studies have thoroughly evaluated the appropriateness of housekeeping genes for RT-qPCR data normalization in Syrian hamsters. The four algorithms (NormFinder, geNorm, Bestkeeper, Delta Ct method) identified slightly different stability of the candidate genes in each tissue which may reflect the differences between the input data and the mathematical models used by each software program. Further integration of these data in RefFinder yielded comprehensive gene ranking from the most stable to the least stable genes in each of the four tissues. The integrated ranking generated through RefFinder differed from the rankings generated from each of the individual analyses. GeNorm identified Actb/Rps18, Actb/Tubb, Rps18/Rpl13a and B2m/Ywhaz as the reference genes to be used for proper normalization of transcripts in lung, liver, spleen and pancreas respectively in GSHs. When additional data from the other program were incorporated through RefFinder, this shifted to Actb, Hprt1, RpI13a and Tubb in the lung, liver, spleen and pancreas respectively. A previous study using only GeNorm identified Cdkn1 and G3pdh as the most appropriate housekeeping genes in aortic tissue of Syrian hamsters (Rueda-Martinez et al., 2016). Given the differences we found between GeNorm and RefFinder, it may be appropriate to reevaluate these findings with additional methods. Indeed, the selection and validation of housekeeping genes in multiple software programs is necessary before assuming the consistency of their application to different tissues.

We identified Actb, Rpl13a, Hprt1 and Tubb as the most stably expressed genes in the lung, spleen, liver and pancreas respectively. We recommend using these genes in future studies investigating gene expression in hamsters using RT-qPCR. We have provided the primer sequences that can be used to amplify these genes.

This research was supported by the PhD research startup foundation of Henan University of Chinese Medicine (RSBSJJ2018-36) and Henan University of Chinese Medicine (38103009-2020-2). Thank you for the anonymous reviewers and the editor of the journal for their valuable comments.