The Reference Gene Selection to Study PRNP Gene Expression in Sheep

The PRNP gene is connected to scrapie susceptibility in sheep (prion disease). Its polymorphism and expression may influence the occurrence of the disease. In order to study PRNP gene expression level in different ovine tissues, selection of reference genes is needed. Three housekeeping genes (RPL27, RPS29, OAZ1) were chosen for studying PRNP gene expression in ovine brain cortex, midbrain, cerebellum, brain stem, pituitary gland, spleen, liver, skeletal muscle and heart. The primers for gene sequencing were designed based on bovine reference sequences. RPL27 was found to be the most stable reference gene (for brain tissues M=0.322, SDCt=0.486, Stability Value=0.0089; for all tissues M=0.489, SDCt=0.696; Stability Value=0.0093). Regardless of the housekeeping gene, the expression level of PRNP was higher in brain tissues than in other tissues analyzed. A normalization experiment indicated that all candidate reference genes could be used as endogenous controls for studying PRNP mRNA expression in different ovine tissues. However, RPL27 seemed to be the most stable and appropriate for the experiment with many different tissue types and could be used as one of the reference genes for studying gene expression in ovine tissues. Our results confirmed that the PRNP gene is highly expressed in nervous tissue.

Scrapie -a fatal, neurodegenerative prion disease of transmissible spongiform encephalopathies (TSE) -affects sheep and goats.The pathogenic prion protein (PrP Sc ) is thought to be the putative TSE agent.The PRNP gene encodes a physiological, cellular form of prion protein (PrP C ) and its polymorphism influences the scrapie susceptibility and incubation period.According to PRUSINER (1998), the conversion of normal PrP C into the pathological prion PrP Sc causes the prion diseases.PrP Sc , the protease-resistant isoform of cellular PrP C , accumulates in brain and other organs of affected animals leading to tissue degeneration (PRUSINER 1998;BAYLIS & GOLDMANN 2004;TRANULIS et al. 2011).Therefore, PRNP gene expression levels may influence scrapie pathogenesis.The disease has two forms: classical and atypical, which differ in e.g.PrP Sc glycosylation profile and genetic susceptibility.On the basis of PRNP genotype, scientists classified sheep into five classical scrapie risk groups (with increasing susceptibility).By contrast, atypical scrapie was diagnosed in animals with classical scrapie resistant genotypes and genotypes associated with classical scrapie susceptibility did not correlate with atypical scrapie occurrence (BAYLIS & GOLDMANN 2004;TRANULIS et al. 2011).Because of its role in the disease, researchers investigated the PRNP gene and protein expression mainly in nervous tissue (BAYLIS & GOLDMANN 2004; GARCIA-CRE- SPO et al. 2005; 2006).gene and protein expression mainly in nervous tissue (BAYLIS & GOLDMANN 2004; GARCIA-CRE- SPO et al. 2005; 2006).
The real-time quantitative PCR (qPCR) gene expression analysis is a very precise and sensitive method.It allows the detection of small differences in gene transcript levels between cell lines, tissues, samples, etc.However, the method requires data normalization against stably expressed endogenous reference genes to obtain accurate results.The expression level of endogenous controls often depends on many factors e.g.tissue, age, diet, living conditions, environment (BUSTIN et al. 2009;CHAPMAN & WALDENSTRÖM 2015).For gene expression experiments BUSTIN et al. (2009) and CHAPMAN and WALDENSTRÖM (2015) recommended: the selection of appropriate reference genes, careful organization of study groups, collect material, isolate and process RNA under appropriate conditions.
The aim of our experiment was to select and evaluate accurate endogenous controls for studying PRNP gene expression and reveal PRNP mRNA expression levels in different ovine tissues.

Tissue collection and RNA extraction
Tissue fragments were collected directly after regular slaughter (intended for human consumption) of five adult healthy Polish Merino sheep (females, purebred, 101-125 months old) from the same flock.The tissues were immediately stored in liquid nitrogen.Total RNA was isolated from brain cortex, midbrain, cerebellum, brain stem (obex), pituitary gland, spleen, liver, skeletal muscle and heart (apex cordis) by using the SV Total RNA Isolation System (Promega).The RNA quan-tity and purity was evaluated by using a NanoDrop spectrophotometer (A260/280 and A260/230 ratios).The RNA quality was proved by gel electrophoresis (1% agarose).Next, concentration of RNA was normalized and reverse transcribed (1000 ng per sample; High-Capacity cDNA Reverse Transcription Kit; Applied Biosystems).cDNA from the same tissue was poolled into one sample and a cDNA 5-point dilution series (1:10) was prepared separately for every tissue analyzed.

cDNA synthesis and quantitative real-time PCR
Relative quantification (ÄÄCt method) and qPCR efficiency estimation (Standard Curve) were performed with standard protocols in the Ste-pOnePlus Real-time PCR System (Applied Biosystems), in 15 µl reaction volume containing: 7.5 µl TaqMan Fast Advanced Master Mix, 900 nM of each primer, 250 nM of each TaqMan probe, 1 µl cDNA and nuclease-free water.The standard curve method was performed separately for all genes studied.After analysis, the most diluted samples (5 th dilution) were rejected because of many outliers in replicate groups.In the comparative ÄÄCt method, the PRNP gene probe was multiplexed with a candidate endogenous control in three separate reactions for OAZ1, RPL27 and RPS29.All reactions were carried out in triplicate.Skeletal muscle was used as a reference to calculate the relative PRNP gene expression (RQ) among tissues.

Data analysis
The reaction efficiency was estimated in

Results and Discussion
DE JONGE et al. (2007) placed OAZ1, RPL27 and RPS29 among the most stably expressed genes in human meta-analysis.We tested these candidate internal control genes along with the PRNP gene.The PCR efficiency (E) values were estimated for all genes and tissues separately and also for all brain tissues together on the basis of the Standard Curve method with a cDNA 4-point dilution series.Although all E values were close to 2, they differed between genes and tissues.The lowest PCR efficiency was obtained for RPS29 in midbrain (1.880), and the highest for PRNP in heart and OAZ1 in brain stem (2.013).The correlation coefficient values were between 0.985 and 1 (Table 2).Mean efficiencies were used for PRNP gene expression studies (ÄÄCt).The Ct Means ranged from 21.68 in brain cortex to 25.49 in heart for OAZ1, from 21.57 in brain cortex to 24.81 in heart for RPL27 and from 20.01 in skeletal muscle to 23.25 in heart for RPS29 (data not shown).
The geNorm algorithm calculates the gene expression stability measure (M) for internal control genes (VANDESOMPELE et al. 2002).The reference gene is stable if its M value is lower than 1.5.For all candidate endogenous controls in both groups: brain tissues and all tissues together (Table 3), the M value was lower than 1.5.Although the BestKeeper algorithm uses Ct Standard Deviation (SD Ct ) and Correlation Coefficient (r) for the gene expression stability calculation (PFAFFL et al. 2004), the results obtained for OAZ1, RPL27 and RPS29 were the same -all genes were stable (with SD Ct 1 and high r).NormFinder measures the stability value (combination of intra-and intergroup variation) -the lower the stability value, the greater the stability of the gene (ANDERSEN et al. 2004).Stability values in NormFinder were very low for all studied genes (Table 3).However, according to NormFinder, geNorm and BestKeeper, the most stable reference gene for all tissues including brain tissue was RPL27 with the lowest M, the lowest SD Ct and the lowest Stability Value.As for the other two genes, on the basis of all algorithms RPS29 stability was better than OAZ1 in brain tissues (except for OAZ1 r value for brain tissues in BestKeeper) and two algorithms (geNorm and NormFinder) showed that OAZ1 stability was better than RPS29 in all tissues (Table 3).OAZ1 and RPL27 have not been tested in sheep gene expression normalization studies before.However,  OAZ1, RPL27 and RPS29 have not been as popular as ACTB, GAPDH or 18S RNA for use as endogenous control genes.However, our findings showed that these three genes could be a good choice for gene expression normalization studies in sheep.
Furthermore, PRNP mRNA expression pattern in ovine tissues was investigated in relation to three candidate housekeeping genes in multiplexed reactions: PRNP + OAZ1, PRNP + RPL27 and PRNP + RPS29 (Fig. 1).As expected, the PRNP mRNA abundance was higher in nervous tissue and pituitary gland than in other tissues analyzed, regardless of the endogenous control.HAN et al. (2006) obtained similar results, but their experiment was conducted with SYBR Green instead of TaqMan probes.Research on PRNP gene expression did not reveal such differences between brain tissues and spleen (GARCIA-CRESPO et al. 2005).In fact, the mean PRNP mRNA expression in spleen was higher than in cerebrum and cerebellum.In our study, in analyzed tissues except brain tissue, the highest PRNP gene expression level was observed in heart.HAN et al. (2006) found that the PRNP gene expression level in heart was lower than in spleen, but higher than in liver.It is not clear, which part of the heart HAN et al. (2006) investigated.
The results of PRNP gene expression analysis varied depending on endogenous control used (Friedman's ANOVA: Chi^2 ANOVA = 6.250 (N=9, df=2), p=0.04394).When normalized with OAZ1, the highest PRNP mRNA expression was observed in brain stem (followed by midbrain, brain cortex and cerebellum) and the lowest in   (2006) showed that the highest mRNA abundance was observed in obex and neocortex followed by cerebellum, spinal cord, hippocampi, conarium and thalamus.Some discrepancies between our results and the mentioned studies could be explained by differences in chosen housekeeping genes, sheep age and breed, number of samples and, in some cases, also in method.GOSSNER et al. (2009) showed that PRNP gene expression might be influenced by PRNP genotype.They studied VRQ/VRQ, ARR/VRQ and ARR/ARR sheep and found that in heterozygotes the PRNP levels were higher than in homozygotes.However, in previous studies GARCIA- CRESPO et al. (2005) did not confirm the hypothesis that PRNP mRNA levels vary between genotypes.Animals used in our study had ALRR/ALRR, ALRR/AFRQ and ALRR/ALRQ genotypes.Because of the small number of animals, we couldn't check statistically whether these genotypes influenced PRNP gene expression level in our study.Quite interesting results were obtained in our experiment for pituitary gland, where PRNP mRNA abundance was almost as high as in cerebellum.To our knowledge, PRNP mRNA expression has not been reported to date in ovine pituitary gland.In cattle, PRNP expression in pituitary gland was lower than in nervous tissues, spleen, liver and had the same level as in muscle (TICHOPAD et al. 2003).Further research is needed to reveal the role of PRNP in the pituitary gland.

Conclusions
Our results confirmed that the PRNP gene, which plays an important role in scrapie pathogenesis, is highly expressed in nervous tissue and showed that the PRNP mRNA level was also high in pituitary gland.According to geNorm, Best-Keeper and NormFinder analysis, the RPL27 housekeeping gene was the most stable among the three tested candidate endogenous controls (RPL27, OAZ1, RPS29) for studying PRNP mRNA abundance in ovine tissues.Therefore, RPL27 can be considered as one of the reference genes for qPCR normalization studies in ovine tissues.

Table 1
Primers and probes.A -primers used for reference gene sequencing; B -reference gene primers and TaqMan MGB probes for real-time PCR; C -PRNP primers and TaqMan MGB probes for real-time PCR

Table 3
(WANG et al. 2010)t al. 2016lculations by geNorm, BestKeeper and NormFinder; M -stability measure (geNorm); SD Ct -Ct Standard Deviation (BestKeeper); r -Correlation Coefficient (BestKeeper); SV -Stability Value (NormFinder); bold font -the optimal value according to the calculation tools have been analyzed in other species.The RPL27 gene was the second best in DE JONGE's human "top 15 candidate housekeeping genes" list (2007).ROPKA-MOLIK et al. (2012)confirmed that OAZ1 is the best internal control for studying gene expression in porcine uterus and ovary and RPL27 is the most stable in porcine oviduct.OAZ1 and RPL27 were the most suitable reference genes in porcine adipose tissue(PIÓRKOWSKA et al. 2011).OAZ1, RPS29 and RPL27 had the highest stability in porcine stomach(OCZKOWICZ et al. 2010).Conversely, RPS29 had the lowest stability in a preterm lamb model with lung injury(PEREIRA-FANTINI et al. 2016) and was relatively unstable in epithelial and nonepithelial cells of mouse small intestine(WANG et al. 2010).For studying PRNP relative gene expression in ovine cerebrum, cerebellum, obex, spleen, terminal ileum and mesenteric lymph node, GARCIA-CRESPO et al.